Royal Jelly: what the bee biology shows, what human trials show, and the allergy risk that matters

Summary

Royal jelly is a bee secretion whose most famous property is real insect biology that does not transfer to people, and it is one of the very few dietary supplements with a documented human death from allergic reaction, so the honest read is that the human evidence is strongest on the harm side and weakest on the benefit side.

Why Emerging

This entry sits at Emerging overall, with the sub-areas splitting sharply — and the split is the point.

Why not the tier above (Moderate). A Moderate rating would require a consistent body of human trial evidence supporting some real-world outcome. That does not exist. The best-designed positive trial has 42 participants, two significant endpoints out of roughly two dozen tested, no described multiplicity correction, an effect on hot flushes that appeared and then vanished within the trial, and direct manufacturer funding with five author-employees. The largest pooled samples in the literature run to a few hundred people. Where heterogeneity is reported it reaches 98 percent, which makes pooled estimates near-uninterpretable regardless of the certainty label attached. The single trial with a hard clinical endpoint was null. And no health claim is authorised in the EU.

Why not the tier below (Experimental). Experimental would imply the human evidence is essentially absent or mechanistic-only. That understates it. Randomised, double-blind, placebo-controlled trials do exist; risk of bias in the individual trials is rated mostly low; two independent meta-analyses report a total-cholesterol reduction; oxidative-stress markers move in a coherent direction across independent trials; and there is a positive, pre-specified signal on anxiety and back pain. That is more than nothing, even if it is much less than the marketing implies.

Per sub-area:
• Composition and 10-HDA specification: Foundational. Analytical chemistry with an international standard. Not contested.
• Caste biology as bee biology: Foundational. The feeding regime and the queen-versus-worker phenotype are settled entomology.
• The royalactin mechanism: Experimental. One single-author paper, a published failure to reproduce in the same journal backed by a depletion-and-enrichment experiment, and an unresolved reply. Journal prestige does not raise this.
• The absent-inhibitor (p-coumaric acid) explanation: Emerging. One laboratory, in-vitro rearing, a partial phenotype, no independent replication — but a real published experiment with a coherent mechanism.
• Chromatin and multi-mechanism accounts: Experimental. Verified at title level only; cited solely to establish that the mechanism question is open.
• Species transfer and digestion: Moderate — for the negative. One well-designed study of the precondition, and it found the precondition unmet in the intestinal phase. Moderate confidence that intact royalactin does not act systemically in humans; that is a moderate-confidence negative, not a benefit rating.
• Menopausal symptoms: Emerging. The best human indication, and still small, short, single-centre, multiplicity-vulnerable and industry-funded.
• Lipids: Emerging. Two meta-analyses, a small total-cholesterol effect with substantial heterogeneity, no LDL movement, and a dose-response running the wrong way on HDL and triglycerides. Full texts unread here.
• Inflammation and oxidative markers: Emerging. Real randomised trials, real GRADE assessment, and heterogeneity that undercuts the pooled numbers. Markers only.
• Glycaemic control: Moderate, for the null. A tight, zero-heterogeneity null on fasting glucose. Moderate confidence in no effect, limited by a small pooled sample.
• Wound healing: Emerging, negative. One placebo-controlled human trial with a hard endpoint, underpowered, null.
• Fatigue and quality of life: Emerging. Two small trials in oncology populations, full texts unread, comparator problems in one.
• Testosterone, fertility, cognition: Experimental. Rodent damage-rescue models and one three-ingredient combination trial. No human evidence for royal jelly alone.
• Allergy and anaphylaxis: Moderate. Multiple independent human case reports and series, a documented fatality, a population sensitisation survey with skin-test data, and laboratory cross-reactivity work. This is the strongest human evidence in the entry, and it is on the harm side.
• Incidence of severe reaction: not established. No tier is assigned because no denominator exists.

The confidence field reads "moderate on harms, low on benefits" precisely because a single tier cannot express that inversion.

Practical takeaway

If you have asthma, eczema, hay fever, or any diagnosed atopy: don't. This is the strongest practical statement in the entry. The documented severe reactions, including the fatal case, concentrate in exactly this group, the reaction can be bronchospasm rather than a rash, and there is no demonstrated benefit on the other side of the ledger that would make the trade sensible. Prior sensitisation to house dust mite may mean the first dose is not a safe trial dose.

If you are not atopic and want to try it anyway, the honest framing is that you are running a personal experiment on a substance with no authorised health claim and a thin trial literature. Some things that follow from that:
• Have a reason and an endpoint. The only indication with any positive placebo-controlled human signal is menopausal symptoms, and even there the trial was small, industry-run, and produced two hits across roughly two dozen endpoints. If you are taking it for energy, testosterone, cognition or longevity, the human evidence for that specific use is absent, not merely weak.
• First dose, small, at home, with someone around, not before exercise or bed. Reactions in the case literature began within minutes. This is standard first-exposure caution for a known potent allergen.
• Doses in the trials ranged from a few hundred milligrams to about 3 grams per day. The menopause trial used 800 mg per day of enzyme-treated royal jelly for 12 weeks. In the lipid meta-analysis the total-cholesterol subgroup effect appeared at 3000 mg per day and above — but the same review's dose-response analysis found HDL falling further and triglycerides rising further at higher doses, so more is not straightforwardly better. There is no established minimum effective dose for anything, because nothing is established.
• Take the label's 10-HDA number for what it is. It is a purity and authenticity marker, not a potency measure. A freeze-dried powder shows a higher percentage than fresh jelly because the water was removed. No source links 10-HDA content to any human outcome. The international standard that defines the marker is voluntary, and adulteration with cheaper material is documented.
• Give it a defined trial and then stop. Twelve weeks is the length of the best trial. If nothing has changed that you can name, the reasonable conclusion is that it is not doing anything for you, because the population-level evidence gives you no reason to expect a slow build.

Tell your prescriber if you are on anything. Specifically: warfarin (one case report of a possible interaction with elevated INR and bleeding), any glucose-lowering drug (royal jelly has been flagged for possible glucose-lowering at high doses, which is a caution rather than a benefit), and blood pressure medication.

Where royal jelly sits against alternatives. For menopausal symptoms, the evidence base for hormone therapy is on a different scale entirely; see menopausal_hormone_therapy_evidence. For lipids, glycaemic markers, and inflammation, the diet and training levers described in whole_food_emphasis are better evidenced by orders of magnitude and cost less. Royal jelly is not competing with nothing.

Evidence detail

Why This Entry Exists

Royal jelly occupies an unusual position in the supplement aisle. Almost every other supplement is sold on human evidence that is weak but at least pointed in the right direction, with harms that are theoretical. Royal jelly is the reverse. The evidence for benefit in humans is small, short, heterogeneous and mostly measured in blood markers rather than in anything a person feels. The evidence for harm in humans includes case reports, a documented fatality in a person with pre-existing asthma, and a government-mandated warning label across Australia and New Zealand. That inversion is the single most useful thing to know about this product, and it is almost never how it is presented.

The reason people buy royal jelly is not any of the human trials. It is the queen bee. The story goes that a bee larva fed royal jelly becomes a queen — larger, fertile, and living years where her sisters live weeks — and the implication, never quite stated out loud, is that something that dramatic must do something for you. The bee facts underneath that story are true and genuinely remarkable. The inference is a species error, and it is doing all the selling. Nobody has to make a human claim, because the buyer makes it for them, which is also why the marketing is hard to fact-check at the level of any individual sentence.

There is a second reason the entry exists. The mechanism most often cited to bridge bee to human — a royal jelly protein called royalactin — was published in Nature and then formally challenged in Nature by researchers who could not reproduce it and who ran a further experiment against it. The dispute is unresolved, both sides published, and a serious competing explanation says queen fate may depend on what royal jelly lacks rather than what it contains. Presenting royalactin as settled science, in either direction, misrepresents the primary literature.

What bad advice this protects against, in all directions:
• "Royal jelly makes queens, so imagine what it does for you." Insect caste determination operates on a larva during a developmental window, through juvenile hormone and insect signalling pathways that humans do not possess. It predicts nothing about an adult mammal taking a capsule, and the mechanism is contested even within bee research.
• "Only queens get royal jelly." Every female larva is fed royal jelly for about three days. What distinguishes a queen is that she receives it exclusively and in abundance, never anything else. That correction matters, because royal jelly alone is demonstrably not sufficient to make a queen.
• "It's a natural food, so it's safe." Royal jelly has caused anaphylaxis and life-threatening asthma, and a fatal case is in the medical literature. Australia and New Zealand require a warning statement on food containing it. The risk concentrates in people with asthma, eczema, hay fever or other atopy — which overlaps heavily with the population drawn to "natural immune support".
• "It's dangerous, avoid it." The opposite overreach. For a person without asthma or atopy, the severe-reaction literature is scattered case reports with no denominator, and the trials that exist reported no adverse events. Both facts ship together.
• "Clinically proven to lower cholesterol / reduce inflammation / boost testosterone." Total cholesterol moves modestly in pooled trials while LDL does not, which is mechanistically incoherent — and in the one dose-response analysis, higher doses pushed HDL down and triglycerides up. Inflammation markers move with heterogeneity so extreme the pooled estimate is barely interpretable. The testosterone claim rests overwhelmingly on rodents rescued from chemotherapy drugs.
• "No side effects were reported in the clinical trials." True, and misleading. Those trials enrolled twenty to a hundred people for four to twelve weeks. They could not possibly detect the one serious risk that actually exists.
• "Regulators approve it." No health claim for royal jelly is authorised in the EU. It is sold as a food supplement in the EU, US and UK, meaning no regulator reviewed the specific product for safety or efficacy before it reached the shelf.

What this entry OWNS: the royal jelly evidence base specifically — its composition and quality markers, the caste-determination story and its scientific dispute, the species-transfer question, the human trial literature by indication, the allergy and anaphylaxis risk profile, and the regulatory posture across jurisdictions.

What this entry DEFERS: general supplement dosing, form and bioavailability reasoning to supplement_form_elemental_dose_and_bioavailability; menopause management as a clinical question to menopausal_hormone_therapy_evidence; immune function as a system to immune_function_cross_pillar_optimisation; antioxidant-marker reasoning and why marker movement is weak evidence to polyphenols_evidence_and_the_antioxidant_myth; atopic disease management to hay_fever_allergic_rhinitis_sleep_protocol and eczema_skin_barrier_lifestyle; and the general case for food over isolated compounds to whole_food_emphasis.

Evidence

Read the tiers, not the thesis. This is a topic where a single vivid fact does most of the persuasive work and the actual claim structure is thin underneath it. Every numbered claim below is labelled HUMAN or preclinical, because the gap between those two labels is where this entire topic lives.

Sub-area 1: What royal jelly is (compositional facts)

1. Royal jelly is roughly 60 to 70 percent water, 9 to 18 percent protein, and 12 to 20 percent total sugars, with a lipid fraction. Foundational. Preclinical — this is analytical chemistry, not a claim about any organism. A 2024 compositional and quality study of royal jelly samples reports the moisture, protein and total-carbohydrate ranges, and measured total amino acids at roughly 11 to 13 grams per 100 grams as-is. The lipid percentage commonly quoted online was not confirmed against a source here and is stated qualitatively rather than numerically. What matters practically: composition varies with what the beekeeper feeds the colony — the same study found sugar-fed and honey-fed colonies produced measurably different sugar profiles. Any single "royal jelly contains X" number is a range, not a constant. Note also the dose arithmetic: a one-gram supplement dose supplies about a tenth of a gram of protein, which is nutritionally trivial. (Nutrient Composition and Quality Assessment of Royal Jelly Samples Relative to Feed Supplements, PMC11203284, 2024.) Cui bono: vendors benefit from the "complete amino acid profile" framing, which is technically true and practically irrelevant at supplement doses. Debunkers benefit from the identical numbers read the other way. The dose settles it, not the composition.

2. 10-HDA is royal jelly's marker fatty acid, present at roughly 1.4 to 3 percent, and is the international authenticity standard. Foundational. Preclinical — an analytical specification. ISO 12824:2016 (Royal jelly — Specifications) sets a minimum of 1.4 grams per 100 grams for fresh royal jelly and, per multiple concurring secondary sources, 3.5 grams per 100 grams for freeze-dried. The 2024 compositional study measured 1.9 to 2.6 grams per 100 grams in its samples. 10-HDA works as a marker because royal jelly is reported to be its only natural source, so adulteration dilutes it. Two things follow that vendors routinely elide. First, this is a purity specification, not a potency measure — no source located here links 10-HDA content to any human health outcome. Second, a higher percentage on a freeze-dried powder label reflects removed water, not greater potency. ISO standards are also voluntary unless a jurisdiction writes them into law, and none of the EU, US or UK does. Published analyses have found a proportion of commercial samples failing the standard's limits; the specific failure rate circulating could not be tied to a retrievable source and is deliberately not given here. Cui bono: a certificate of analysis reads as pharmaceutical rigour and justifies a price premium, so premium producers benefit heavily from a strict marker standard. Analytical laboratories selling authentication services benefit from adulteration being perceived as widespread. The neutral point that survives both incentives is that a consumer cannot tell from the label what is in the jar.

Sub-area 2: The queen bee story (all insect biology)

3. All female honeybee larvae are fed royal jelly for about three days; queen-destined larvae are fed it exclusively and in abundance. Foundational. Preclinical (insect). Mao, Schuler and Berenbaum state directly that nurse bees feed queen-destined larvae exclusively royal jelly, while worker-destined larvae receive royal jelly for three days and subsequently jelly to which honey and beebread are added. The popular shorthand that only queens get royal jelly is wrong in a way that matters: royal jelly by itself is not sufficient to produce a queen, since every worker also received it. Queens live years against a worker's roughly thirty-five days in the foraging season, and are the colony's sole reproductive female. (Mao W, Schuler MA, Berenbaum MR, Science Advances 2015, doi:10.1126/sciadv.1500795, PMID 26601244; queen-versus-worker lifespan comparison, PMC9506344, cited by title.) Cui bono: this is the most valuable asset the industry owns — a true, vivid, verifiable fact that does all the persuasive work by implication, so nobody has to make a human claim. The same fact stated in full serves the sceptic: the queen is long-lived because she is a reproductive queen, and the three-day universal feeding shows royal jelly alone is not the cause.

4. A 2011 Nature paper reported that one royal jelly protein, named royalactin, induces queen development and queen-like phenotypes in fruit flies. Experimental. Preclinical (insect). Masaki Kamakura reported that a 57-kilodalton royal jelly protein (major royal jelly protein 1), which he designated royalactin, induces differentiation of honeybee larvae into queens — increasing body size and ovary development and shortening developmental time — and that royalactin also produced queen-like phenotypes in Drosophila melanogaster through epidermal growth factor receptor signalling. This is a single-author Nature paper, which is relevant to how much independent internal replication it carried. Tiering it as Experimental rather than higher is deliberate: journal prestige is not evidence quality, and this specific result has a published failure to reproduce in the same journal. (Kamakura M, Nature 473:478–483, 2011, doi:10.1038/nature10093, PMID 21516106.) Cui bono: overwhelmingly the supplement industry — royalactin is the mechanism story that converts a bee fact into a human product, and "Nature, 2011" is the credential that carries it.

5. A 2016 Nature paper reported failure to reproduce royalactin's queen-inducing role, and went further than a bare replication attempt; the dispute is unresolved. Moderate. Preclinical (insect). Buttstedt, Ihling, Pietzsch and Moritz published Royalactin is not a royal making of a queen as a Brief Communications Arising — the format Nature uses for a formal challenge to a paper it previously published. They reported unsuccessful attempts to repeat some of the original royalactin experiments, and they also ran a designed test rather than a bare replication: larval food either depleted of royalactin or enriched with it produced no difference in queen caste determination, and they extended the exclusion to MRJP2, MRJP3 and MRJP5 as single key drivers. Kamakura published a reply in the same issue. Both parties are on the record in the same journal and neither side has been retracted. A failed replication is not proof of absence any more than a single positive is proof of presence; in-vitro honeybee larval rearing is technically demanding and outcome-sensitive to diet, humidity and handling, which is exactly the ground such exchanges land on. The honest summary is: contested, unresolved — while noting that a depletion-and-enrichment experiment is a stronger form of challenge than a non-replication alone. (Buttstedt A, Ihling CH, Pietzsch M, Moritz RFA, Nature 537(7621):E10–E12, 2016, doi:10.1038/nature19349, PMID 27652566; Kamakura reply, doi:10.1038/nature19350.) Cui bono: nobody sells "royalactin does nothing", so the sceptical side has thin commercial motive — that asymmetry is itself informative. An entire supplement category rests on the 2011 result and none of it rests on the 2016 one.

6. A competing explanation inverts the story: queen fate may depend on what royal jelly LACKS. Emerging. Preclinical (insect). Mao, Schuler and Berenbaum report that p-coumaric acid, a plant phenolic, is ubiquitous in honey and beebread but absent from royal jelly; that it differentially regulates genes involved in caste determination; and — the key experiment — that rearing larvae in vitro on a royal jelly diet with p-coumaric acid added produces adults with reduced ovary development. Their interpretation is that exclusive royal jelly feeding may have evolved alongside exploitation of plant chemistry to suppress worker reproduction. On this account royal jelly is not special for what it contains but for what it omits. Tiered Emerging, not higher: one laboratory, an in-vitro rearing system, and the measured outcome is reduced ovary development rather than the full queen-versus-worker phenotype. It shows p-coumaric acid is capable of shifting caste outcomes; it does not establish that it is the determinant in a real hive. The two hypotheses are also not mutually exclusive. (Mao W, Schuler MA, Berenbaum MR, Science Advances 2015, PMID 26601244.) Cui bono: this finding is commercially inconvenient and is accordingly almost never mentioned on product pages — the selective silence is itself informative. It has no commercial sponsor of its own.

7. At least three distinct mechanisms have been proposed for caste determination, which is the strongest evidence none is settled. Experimental. Preclinical (insect). Alongside royalactin and the absent-inhibitor account, a chromatin-level explanation has been proposed — a paper indexed as PMC3059907 is titled to the effect that histone deacetylase inhibitor activity in royal jelly might facilitate caste switching, and a Communications Biology paper is titled to argue against any single master-inducer account. Both were verified only at title level during research; neither was read, and no author is named for either here. They are cited for one purpose only: to establish that the mechanism question has multiple live candidates. No sentence in this entry rests on their contents. Cui bono: the epigenetics framing is the one most attractive to supplement marketing ("royal jelly changes gene expression"), which is a reason for more citation caution here, not less.

Sub-area 3: The species-transfer question (this is the spine of the entry)

8. Royal jelly proteins are rapidly destroyed by intestinal digestion, and no evidence was found that intact functional protein reaches human circulation. Moderate. HUMAN (in vitro) — simulated human gastrointestinal digestion using human digestive enzymes, not a human trial; do not read the label as more than it says. Mureşan, Schierhorn and Buttstedt examined the fate of major royal jelly proteins through simulated human gastrointestinal digestion. The proteins vary in resistance to pepsin, with MRJP2 the most stable and still detectable full-length after twenty-four hours of gastric-phase conditions. In the intestinal phase with trypsin and chymotrypsin, they are rapidly digested, MRJP2 again lasting longest at about forty minutes. The authors' own conclusion is the careful one: the proteins can reach the small intestine full-length but would then have to be absorbed quickly if full-length proteins are to have any biological activity. That is a statement of an unmet precondition, not a demonstration. No study showing intact royal jelly protein in human blood was located, and no human study of royalactin as such was located. Absence of located evidence is not proof of absence — but the burden sits with the claimant. (Mureşan CI, Schierhorn A, Buttstedt A, J Agric Food Chem 66(16):4164–4170, 2018, doi:10.1021/acs.jafc.8b00961, PMID 29629561.) Cui bono: this is the industry's quiet spot. The whole royalactin sales story requires an intact functional protein reaching human tissue, and the one paper that examined the precondition found it unmet in the intestine. Note the honest counterweight: digestion products are a separate research line — royal jelly hydrolysates and blood-pressure-relevant peptides — so "the protein is digested" disproves royalactin-as-mechanism, not royal jelly activity in general.

9. A specific rhetorical move to watch: MRJP2's stomach survival is cited as though it were whole-body survival. Moderate. HUMAN (in vitro). Vendors sometimes cite the pepsin resistance finding as evidence that royal jelly proteins "survive digestion". That is true of the stomach and false of the intestine, in the same paper, a few sentences apart. The unusual gastric stability buys minutes to hours; it does not deliver a functional signalling protein into the bloodstream. (Same source as claim 8.)

Sub-area 4: Human trials, by indication

10. Menopausal symptoms: the strongest indication, and still weak — and the flagship trial was run by a bee company. Emerging. HUMAN. Asama and colleagues randomised 42 healthy Japanese postmenopausal women, 21 per arm, to 800 mg per day of enzyme-treated royal jelly or dextrin for 12 weeks, with a four-week follow-up, scoring roughly two dozen menopausal symptoms. Two reached significance: anxiety (P=0.046) and backache or low back pain (P=0.040). Hot flushes improved at eight weeks (P=0.026) but the effect was gone by twelve (P=0.363), and no differences remained four weeks after stopping. No side effects were reported. Two hits out of two dozen endpoints in 42 people, with no multiplicity correction described, is the classic shape of a false-positive-prone design; a benefit that appears at eight weeks and vanishes at twelve looks more like noise than like a dose-duration effect. The study was funded by Yamada Bee Company, five authors were company employees, and a further author took advisory fees from the company — all disclosed, to their credit. Two further trials exist in this indication (an Iranian trial reported at roughly 200 women on 1000 mg per day for 8 weeks, and a genitourinary-syndrome trial) but were not retrieved and are not relied on here. Note also that the product tested was enzyme-treated royal jelly, which does not automatically license claims for a generic supplement. (Asama T, Matsuzaki H, Fukushima S, Tatefuji T, Hashimoto K, Takeda T, Evid Based Complement Alternat Med 2018;2018:4868412, PMC5949161.) Cui bono: menopause is the one indication with any positive placebo-controlled human signal, so it is cited hardest — and the citation almost never mentions the funding or the endpoint count. Against reflexive dismissal: this was genuinely randomised, double-blind and placebo-controlled, and anxiety and back pain were pre-specified primary outcomes. It is the best human data this literature has.

11. Glycaemic control: essentially null. Moderate confidence in the negative. HUMAN. Mahboobi, Jafarnejad and Eftekhari pooled five randomised trials. Fasting plasma glucose changed by −0.95 mg/dL (95% CI −5.83 to 3.87, P=0.69) with I²=0 percent across 205 participants — a tight, consistent null. HbA1c changed by −0.32 (95% CI −0.87 to 0.23, P=0.25) with I²=69 percent across 130 participants. The funnel plot showed no publication bias. The authors' stated conclusion is that royal jelly did not beneficially affect markers of glycaemia, and they explicitly flag methodology problems and uncontrolled confounders including diet. The tier reflects moderate confidence in the negative, not a benefit — with a pooled sample that small, a modest effect cannot be excluded, only a large one. A later review reported a fasting-glucose reduction in a subgroup of longer trials in non-healthy populations; that full text was not retrieved, and subgroup effects inside an overall null are hypothesis-generating at best. (Mahboobi S, Jafarnejad S, Eftekhari MH, Complement Ther Med 2019;44:235–241, PMID 31126561.) Cui bono: sellers cite the individual positive Iranian diabetes trials and never the meta-analysis whose title says royal jelly does not improve markers of glycaemia. That bluntness makes it the most useful citation a sceptic has. Equally, "proven not to work" overstates a 205-person pool.

12. Lipids: a modest total-cholesterol signal with no LDL movement, which is mechanistically incoherent — and a dose-response that runs the wrong way. Emerging. HUMAN. Two meta-analyses exist. An earlier one pooling six trials reported a reduction in total cholesterol with no significant change in triglycerides, LDL or HDL. A later GRADE-assessed dose-response review reported total cholesterol lower by 8.75 mg/dL (95% CI −15.50 to −1.99, P=0.01) with substantial heterogeneity (I²=68.5 percent), and reported the effect as stronger in unhealthy participants, in trials of eight weeks or longer, and at doses of 3000 mg per day or more. Neither the overall pooled analysis nor the subgroups showed a significant effect on triglycerides, LDL or HDL. The finding that appears on no product page sits in the same abstract: the dose-response analysis reported a greater reduction in HDL and a greater elevation in triglycerides at higher doses — the two lipid movements a person would not want. Neither full text was retrievable during research (publisher returned an access error), so these figures come from abstract-level retrieval, and no author names are attached to them here. Clinically: the effect is roughly four percent of a typical baseline total cholesterol, the confidence interval approaches zero at one end, a total-cholesterol effect with no LDL effect — LDL being the causal lipid — is a weak and internally odd finding, and where dose does appear to matter it moves two markers unfavourably. (Royal jelly is an effective and relatively safe alternative approach to blood lipid modulation: a meta-analysis, J Funct Foods 2018; The effects of royal jelly consumption on lipid profile: a GRADE-assessed systematic review and dose-response meta-analysis, Clin Nutr ESPEN 2023 — both cited by title, full texts unread.) Cui bono: "clinically proven to lower cholesterol, confirmed by meta-analysis" is the most commercially useful sentence available here and it survives only by omitting that LDL did not move and that HDL and triglycerides moved the wrong way with dose. Against dismissal: the total-cholesterol signal has been reported more than once and is not obviously zero.

13. Inflammation and oxidative-stress markers: moved, with heterogeneity so extreme the pooled numbers barely mean anything. Emerging. HUMAN. A 2025 systematic review pooled six randomised trials (seven treatment arms, 356 participants — 181 intervention and 175 control — from Iran, Japan and Slovenia). Total antioxidant capacity rose (WMD 0.98, 95% CI 0.24 to 1.71, P=0.009) but with I²=98.5 percent. Malondialdehyde fell (WMD −1.79, 95% CI −3.00 to −0.58, P=0.004) with I²=97.4 percent. High-sensitivity CRP did not change (WMD −0.24, 95% CI −0.60 to 0.10, P=0.17), I²=87.2 percent. GRADE certainty was moderate for the first two and low for CRP; risk of bias was rated low in four studies and moderate in two. Publication-bias tests were non-significant — but with six trials those tests have almost no power, so that is not reassurance. The heterogeneity is the story: I² near 98 percent means these trials are not measuring a common effect in any meaningful sense, and malondialdehyde assays in particular are notoriously non-comparable between laboratories. Every outcome here is a blood marker, not a symptom, a function or an event. (Taheri et al., Effects of royal jelly consumption on inflammation and oxidative stress: a systematic review and meta-analysis of randomized controlled trials, PMC12244952, 2025 — abstract-level retrieval; only the first author was captured.) Cui bono: "clinically proven antioxidant" is technically defensible from this review and tells a buyer nothing about whether they will feel different. The fair counter: the reviewers reported the heterogeneity and the null CRP result rather than burying them, which makes this one of the better-conducted reviews in the literature.

14. Diabetic foot ulcers: the one hard clinical endpoint tested, and it failed. Emerging. HUMAN. Siavash and colleagues ran a double-blind placebo-controlled trial at a hospital in Isfahan, Iran: 25 patients with 64 ulcers, all receiving standard care (offloading, infection control, vascular improvement, debridement), randomised to 5 percent sterile topical royal jelly or placebo, followed for three months or until healing. No significant difference on any healing parameter — depth, length and width reduction, time to complete healing, or incidence of complete healing. The authors state they could not confirm any superiority over placebo. No side effects. This is one of very few royal jelly trials with an outcome a patient would notice, and it is negative. It is also underpowered, single-centre, one concentration, and layered on top of good standard care, so it is "failed to show benefit", not "proven useless". (Siavash M, Shokri S, Haghighi S, Shahtalebi MA, Farajzadehgan Z, Int Wound J, PMID 23566071 — retrieved at abstract and listing level across concordant sources; the author list should be re-checked before it is quoted onward.) Cui bono: apitherapy marketing rides the well-established honey wound-care literature and lets it bleed across to royal jelly, a chemically different substance with its own much thinner evidence.

15. Fatigue and quality of life: two small trials in sick populations, neither read in full. Emerging. HUMAN. Two trials surfaced. One in 52 cancer patients reported that honey plus royal jelly produced less cancer-related fatigue than honey alone — an unusual comparator, arguably a stronger test, but one that cannot isolate the royal jelly contribution. One in 33 patients with advanced renal cell carcinoma on tyrosine kinase inhibitors reported less fatigue and anorexia and fewer drug dose reductions in the royal jelly arm. Neither full text was retrieved; both are reported here as pointers, without author attribution. Fatigue and anorexia are highly placebo-responsive, and "fewer dose reductions" is a clinician-decided outcome vulnerable to unblinding. Both populations were on active oncology treatment, so nothing here transfers to healthy-person energy claims. (Effect of processed honey and royal jelly on cancer-related fatigue: a double-blind randomized clinical trial, PMID 27504161; Oral intake of royal jelly has protective effects against tyrosine kinase inhibitor-induced toxicity in patients with renal cell carcinoma, Medicines 2019;6(1):2.) Cui bono: "clinically shown to reduce fatigue" gets lifted onto general-population energy supplements when the trials were in cancer patients on kinase inhibitors. Against dismissal: keeping patients on their oncology dose would be genuinely valuable if real, and deserves a proper trial.

16. Testosterone, sperm quality and "male vitality": the most aggressively marketed claim and the worst supported. Experimental. Preclinical (rodent) for everything retrievable. The retrievable evidence base is dominated by rodent damage-rescue models: royal jelly and sperm parameters, testosterone and lipid peroxidation in mice treated with oxymetholone (PMID 25050300); protection against bleomycin-induced damage in male rats; reproductive parameters in streptozotocin-diabetic rats. These are animals poisoned with a drug and then partially protected — they say nothing about a healthy man's baseline function. Human material appears to exist in low-visibility journals, but search summaries returned mutually incompatible doses for what may be the same study, so no figure and no citation is given here. No placebo-controlled human trial demonstrating that royal jelly raises testosterone or libido was located. Cui bono: "male vitality" sells at a premium on rodents rescued from chemotherapy drugs. A second layer of overreach sits underneath — the fact that royal jelly makes a bee larva into a fertile queen is insect developmental biology and has no bearing on human gonadal function, but it is doing rhetorical work on almost every fertility marketing page. The counterweight: men with oxidative-stress-associated sperm dysfunction are a plausible population for an antioxidant trial, and nobody appears to have run a proper one.

17. Cognition and neuroprotection: no human evidence for royal jelly alone. Experimental. Preclinical (rodent), with one confounded human trial. The Alzheimer's Drug Discovery Foundation's assessment identifies exactly one relevant human trial, and it tested a combination — 750 mg freeze-dried royal jelly plus 120 mg ginkgo extract plus 150 mg Panax ginseng extract, for four weeks, in 66 people with mild cognitive impairment. Ginkgo and ginseng each have their own cognitive literature; attribution to royal jelly is impossible by design. The assessment states plainly that no adequately designed trial has tested royal jelly alone for cognitive benefit, and that no human study has confirmed the rodent findings on neurogenesis and oxidative stress. A crossover study in middle-aged men reportedly found a reduction in an ageing-associated chemokine but no improvement in cognitive or most psychological outcomes; seen at listing level only, and its direction is negative for the claim. Two further trials are registered, meaning the field itself treats this as open. Circulating descriptions of a twelve-week memory pilot and a postmenopausal mental-flexibility trial could not be traced to any primary source and are not cited here; they may be vendor fabrications. (Alzheimer's Drug Discovery Foundation, Cognitive Vitality assessment: Royal Jelly.) Cui bono: "supports brain health" is a premium claim resting on rodents plus one three-ingredient trial. "It rewires gene expression" is deployed in nootropic marketing as if the organism were interchangeable.

Sub-area 5: Harm (where the human evidence actually is)

18. A fatality attributed to royal jelly-induced asthma is documented in the peer-reviewed literature. Moderate. HUMAN. A case report titled Fatal royal jelly-induced asthma appears in the Medical Journal of Australia. The patient is described here only as a person with pre-existing asthma in Australia: the full text was not read, and secondary sources conflict on age and sex, so no descriptor is given. This is one titled fatal case report. No fatality count is stated — the Australia–New Zealand regulator's label wording uses the plural, but that is a regulator's characterisation of a case literature, not a verified tally. A single case report cannot establish incidence; a death in an asthmatic after allergen exposure is nonetheless a strong causal signal by the standards of anaphylaxis attribution. (Bullock RJ, Rohan A, Straatmans JA, Med J Aust 1994;160(1):44, doi:10.5694/j.1326-5377.1994.tb138207.x, PMID 8271989.) Cui bono: this is the most quotable fact about royal jelly and gets recycled by scare content and by competing supplement categories. In the other direction, a death warning on a premium product is a direct sales suppressant, so the trade has a real interest in it staying obscure.

19. Severe reactions concentrate in people with pre-existing asthma and atopy. Moderate. HUMAN. The Australian case series is the anchor. Thien and colleagues reported royal jelly-induced asthma in 1993, and asthma and anaphylaxis induced by royal jelly in 1996. A further paper on clinical characteristics and immunologic correlations exists (Leung, Thien, Baldo and Czarny, J Allergy Clin Immunol 1995) but its full text returned an access error during research and no finding is attributed to it here. Secondary summaries consistently describe affected patients as having prior mild-to-moderate asthma, allergic rhinitis, eczema or atopic dermatitis. Ascertainment bias runs hard in this direction — atopic people are more likely to see an allergist and be written up — so the direction is well-supported while the strength of the enrichment cannot be read off case series. (Thien FC, Leung R, Plomley R, Weiner J, Czarny D, Med J Aust 1993;159(9):639, doi:10.5694/j.1326-5377.1993.tb138070.x, PMID 8123114; Thien FC et al., Clin Exp Allergy 1996;26(2), PMID 8835130.) Cui bono: naming the at-risk group honestly costs vendors little and serves the reader — it lets a non-atopic person proceed. Vendors prefer the vaguer "may cause allergy in sensitive individuals", which reads as boilerplate and gets ignored by exactly the asthmatic who needs it.

20. Sensitisation is measurable in the population; clinical reaction rate is not. Moderate. HUMAN. Leung and colleagues surveyed 1472 hospital employees in Hong Kong: 461 (31.3 percent) had ever taken royal jelly, and nine subjects reported fourteen adverse reactions in total, comprising urticaria, eczema, rhinitis and acute asthma. Skin-test positivity to pure royal jelly was 13 of 176 questionnaire respondents tested (7.4 percent) and 23 of 300 consecutive asthma-clinic attendees (7.3 percent). All but one of the 36 skin-test-positive subjects were atopic to other common allergens; the association between a positive royal jelly skin test and atopy was OR 33.73 (95% CI 4.51 to 252.11) — direction trustworthy, magnitude not, given that interval. The association between reported adverse reactions and a history of clinical allergy was OR 2.88 (95% CI 0.72 to 11.58), which crosses one and is not statistically significant — note that the paper's own abstract describes this as a positive association, which the confidence interval does not support. The authors' own conclusion is that the relationship between use, a positive skin test and clinical reactions remains to be defined. Sensitisation is not allergy, and a 7.4 percent skin-test rate is not a 7.4 percent reaction rate. (Leung R, Ho A, Chan J, Choy D, Lai CK, Clin Exp Allergy 1997;27(3):333–336, PMID 9088660.) Cui bono: with no incidence figure, "rare" can be asserted indefinitely without contradiction, which serves sellers. Quoting 7.4 percent as though it were a reaction rate would be the mirror-image overreach.

21. The incidence of severe reaction is not established, and that cuts both ways. HUMAN, and the honest statement of a gap. No incidence or prevalence study of severe reaction to royal jelly was located in any country. The word "rare" appears in the Australia–New Zealand warning label; that is regulatory language, not a measured rate, and this entry does not borrow it as one. What can be said: royal jelly is consumed widely, particularly in East Asia and Australia, and the severe-reaction literature consists of scattered case reports and small series rather than any population-scale signal. That is weak evidence that severe events are uncommon in people without asthma or atopy. It is weak in both directions — supplement adverse events are badly under-reported, and the case literature clusters around one Australian allergy community in the 1990s, which suggests reporting-cluster effects as much as true geography. A figure of roughly one in six adults attending asthma clinics having royal-jelly-reactive IgE circulates in review literature; it could not be tied to a named study and is stated here generically, without a citation.

22. Cross-reactivity is real, and the common reassurance about bee venom is wrong. Moderate. HUMAN. A widespread piece of reasoning holds that royal jelly allergens (major royal jelly proteins, a glandular secretion) are a different protein family from venom allergens (phospholipase A2, hyaluronidase, melittin), so venom allergy should not predict royal jelly allergy. That reasoning is wrong as stated: honeybee venom itself contains MRJP8 and MRJP9, so the family is present in venom rather than absent from it. Hata and colleagues tested 30 atopic dermatitis patients with no prior royal jelly exposure: ten of thirty (33 percent) were positive by ELISA, three by skin prick. Inhibition assays showed the royal jelly response was significantly inhibited by house dust mite (three of five patients tested), honeybee venom (four of five) and crab species; German cockroach also cross-reacted. No cross-reactivity with cat dander, Japanese cedar, Japanese mugwort, timothy grass, peanut, buckwheat, shrimp or silk moth. The practically important implication is that prior mite sensitisation may prime a reaction on first exposure. Caveats are heavy: n=30, a selected atopic population, laboratory cross-reactivity rather than demonstrated clinical cross-reactivity, and inhibition sub-analyses resting on five patients. The mite–crab–cockroach pattern also hints the shared epitope may be a pan-arthropod protein rather than an MRJP specifically. Royal jelly shares MRJPs with honey. Propolis reactions are predominantly delayed-type contact dermatitis, a different immunological mechanism. Bee pollen cross-reactivity with royal jelly was not established in anything located. (Hata T, Furusawa-Horie T, Arai Y, Takahashi T, Seishima M, Ichihara K, PLOS ONE 2020;15(6):e0233707, doi:10.1371/journal.pone.0233707.) Cui bono: sellers of bee-product bundles benefit from the categories being treated as separate risk profiles. In the other direction, an overbroad "all bee products are dangerous if you react to one" line is easy to write and would wrongly exclude people.

23. Non-allergic harms exist as isolated case reports: haemorrhagic colitis and one possible warfarin interaction. Emerging. HUMAN. A case report of haemorrhagic colitis associated with royal jelly intake is indexed at PMID 9257239 (no authors named here — the author list was not verified). A warfarin case is documented: Lee and Fermo reported an elevated INR with bleeding in a heavily co-medicated elderly patient, with causality assessed as possible — not probable, not definite. Polypharmacy alone destabilises INR. These are each n=1 and do not support a general interaction claim; they support telling a prescriber. Contact dermatitis is repeatedly listed in secondary reviews; no primary source for it was verified. (Lee NJ, Fermo JD, Pharmacotherapy 2006;26(4):583, doi:10.1592/phco.26.4.583, PMID 16553520.) Cui bono: "no known drug interactions" is technically defensible against an n=1 possible-causality signal and practically misleading to anyone on warfarin. Turning that same case into "royal jelly thins your blood" is the opposite overreach.

24. Occupational sensitisation by inhalation is documented at case-report level. Emerging. HUMAN. A woman who had worked eleven years in a royal jelly processing factory developed wheezing at work that resolved after she left; she was already allergic to honeybee venom, with cross-reactivity confirmed in her case. Two laboratory workers experienced rhinoconjunctivitis and asthmatic symptoms handling royal jelly powder, with allergy confirmed by specific inhalation challenge in one. Full citations were not retrieved (one item is a conference abstract, the weakest publication form) and none is fabricated here. This is a different exposure route and magnitude from consumer ingestion and does not transfer directly — its relevance is mechanistic support that these proteins are potent respiratory sensitisers. A 3.05 percent occupational sensitivity rate that circulates for beekeepers concerns propolis, not royal jelly, and must not be moved across.

25. Children: a paediatric reaction is documented, and the exposure was for an unsubstantiated indication. Emerging. HUMAN. Fantini and colleagues reported a seven-year-old child given royal jelly to promote growth, who had two identical episodes of lip and tongue swelling and palate itch within ten minutes of ingestion, resolving with oral antihistamines both times. The child's sex is not specified in the retrievable record and is deliberately not given here. Prick-to-prick testing with royal jelly extract was positive (6 mm wheal), patch testing gave an immediate positive at twenty minutes, and the child was skin-prick positive to cat dander, various pollens including grass, olive, cypress and Parietaria, and almond. The authors attributed the reaction partly to possible contamination with almond pollen or cross-reactivity with Compositae pollens, which slightly qualifies it as evidence about royal jelly proteins specifically — it may be as much evidence about product contamination. Their conclusion was that atopic individuals may have increased risk of sensitisation to royal jelly. (Fantini P, Delle Donne P, Calogiuri G, et al., Case Rep Med 2014;2014:941248, PMID 24799914, PMC3988729.) Cui bono: no party benefits from a paediatric case report. Note what put the product in the child's mouth — a growth claim no regulator has authorised.

Sub-area 6: Regulation

26. Australia and New Zealand mandate a warning statement naming severe allergic reactions, fatalities and asthma sufferers. Regulatory fact, verified verbatim. Food Standards Australia New Zealand is the bi-national regulator, and the Food Standards Code it administers applies in both countries. It requires the statement: "This product contains royal jelly which has been reported to cause severe allergic reactions and, in rare cases, fatalities, especially in asthma and allergy sufferers." The Code does not permit this wording to be paraphrased. No Food Standards Code clause number is cited here because none was confirmed. This is a labelling mandate, not an epidemiological estimate, and the regulator's phrase "in rare cases" should be read as regulatory language and never converted into a count or a rate. A separate Therapeutic Goods Administration advertising-permission instrument for royal jelly exists in Australia; its wording was not retrieved. Equivalent mandates in the US, UK or EU were not checked, so no claim is made that other countries require this. (Food Standards Australia New Zealand, Warning and advisory statements.) Cui bono: a regulator gets strong public-safety optics from a label warning at near-zero enforcement cost, so mandated wording tends to be maximally cautionary. That is a reason to quote it as a label, not to discount it.

27. The EU has not authorised any royal jelly health claim, on the ground that the substance was not sufficiently characterised. Regulatory fact. The European Food Safety Authority's panel on dietetic products, nutrition and allergies assessed royal jelly claims (IDs 1225, 1226, 1227, 1228, 1230, 1231, 1326, 1328, 1329, 1982, 4696, 4697) under Article 13(1) of Regulation (EC) No 1924/2006. Identified claimed effects include natural defence and immune system, metabolism, and vascular function. The claims were not substantiated, on the ground that the food was not sufficiently characterised — the same objection applied in that batch to ginseng extract and other whole-food constituents. The opinion sits in the fourth Article 13.1 batch published in April 2011, in which 21 of 442 claims were substantiated. The nuance matters and cuts toward royal jelly: "not sufficiently characterised" is a dossier failure, not a finding that royal jelly does nothing — EFSA never reached the efficacy question because it could not establish what the submitted substance was. The full text was behind a paywall during research, so no verbatim conclusion sentence is quoted. (EFSA NDA Panel, EFSA Journal 2011;9(4):2083, doi:10.2903/j.efsa.2011.2083.) Cui bono: EFSA is a public regulator with no commercial stake, so a blanket non-authorisation across twelve claim IDs is a strong negative signal. The honest counterweight: that process was widely criticised for setting a bar almost no botanical could clear — 421 of 442 claims failed — so the failure is not uniquely damning of royal jelly.

28. Royal jelly is a food supplement, not a medicine, in the EU, US and UK — no pre-market review of safety or efficacy. Regulatory fact, assembled from regulatory summaries rather than primary instruments; treat the jurisdictional detail as indicative. In the US it is a dietary ingredient under DSHEA 1994, where the manufacturer, not the regulator, is responsible for safety before marketing, and no medicinal use is approved. In the EU, food supplements fall under Directive 2002/46/EC, with health claims separately requiring authorisation that royal jelly has never obtained. The UK operates a parallel regulated-products regime post-Brexit. The practical consequence is the point: nobody reviewed the specific product on the shelf for safety or efficacy before it got there, which makes the Australia–New Zealand mandated warning the global outlier rather than the norm.

Mechanism

Royal jelly is a glandular secretion produced by nurse honeybees and fed to larvae. Chemically it is mostly water, with a protein fraction dominated by the major royal jelly protein family, a sugar fraction that varies with what the colony is fed, a lipid fraction whose signature component is 10-hydroxy-2-decenoic acid, and assorted vitamins and trace compounds. A commonly repeated attribution of the lipid fraction to the mandibular glands and the protein fraction to the hypopharyngeal glands was not verified during research and is not asserted here.

The mechanism that sells the product is caste determination, and it is worth stating exactly what is and is not known. In the hive, genetically identical female larvae become either short-lived sterile workers or a large, fertile, long-lived queen, and diet is the switch. Three explanations are live. The first is that a specific royal jelly protein actively induces queen development — the royalactin hypothesis, published in 2011 and formally challenged in the same journal in 2016 with a reply from the original author, leaving the question open. The second inverts it: worker diet contains plant-derived compounds, notably p-coumaric acid, that are absent from royal jelly and that suppress ovary development, making the queen's fate a consequence of what she does not eat. The third points at chromatin-level effects on gene expression. These are not mutually exclusive, and a fourth possibility — that sheer quantity and continuous feeding rate do substantial work — has no verified primary source behind it and is raised here only as an acknowledged gap.

Whichever of those is right, the machinery is insect machinery: juvenile hormone, insect developmental windows, insect receptor signalling, and a facultative caste system. Humans have none of it. Even the fruit fly arm of the royalactin paper, often cited as evidence the effect crosses species, crosses from one insect to another insect.

For a human taking a capsule, the chain would have to run: protein survives the stomach, survives the small intestine, is absorbed intact, reaches a tissue, and acts on a receptor that responds to it. The one study that examined the early links found that the proteins are somewhat pepsin-resistant but rapidly destroyed by trypsin and chymotrypsin, with the most stable member gone within about forty minutes of intestinal-phase digestion. Its authors framed rapid absorption as a precondition that would have to hold — an unmet "if". No study showing intact royal jelly protein in human blood was found.

That does not make royal jelly inert, and the distinction matters. Digestion generates peptide fragments, and peptide fragments from royal jelly protein are their own research line, including blood-pressure-relevant peptides studied in rodents. Any plausible human activity almost certainly lives there, or in the small molecules such as 10-HDA, rather than in an intact signalling protein. So the correct mechanistic statement is narrow: the queen-maker mechanism cannot be operating in humans, which is a much smaller claim than royal jelly does nothing.

On the harm side the mechanism is clear and well-characterised, which is part of why the harm evidence is stronger than the benefit evidence. Major royal jelly proteins 1, 2 and 3 are potent IgE-binding allergens. They cross-react in the laboratory with house dust mite, honeybee venom, crab and cockroach — an arthropod pattern — which supplies a plausible route by which a person with no prior royal jelly exposure can be primed to react on first contact. In an asthmatic, an IgE-mediated reaction can present as bronchospasm rather than as hives, which is the mechanism that makes the fatal case plausible rather than freakish.

Risks And Contraindications

Avoid entirely:
• Anyone with asthma. This is the population in which the documented fatal case occurred and in which the severe-reaction case series concentrates. The reaction mechanism in an asthmatic is bronchospasm.
• Anyone with atopic dermatitis, allergic rhinitis, or diagnosed multiple allergies. In the Hong Kong survey, all but one of 36 people with a positive royal jelly skin test were atopic to other common allergens.
• Anyone with known allergy to honeybee venom. The common reassurance that venom and royal jelly allergens are unrelated families is wrong: venom contains major royal jelly proteins 8 and 9, and laboratory inhibition testing showed cross-reactivity in four of five patients tested.
• Children. A paediatric oral allergy case is documented, in a child given royal jelly to promote growth — a claim no regulator has authorised. There is no benefit here worth accepting an unpredictable allergy risk for.
• Pregnancy and breastfeeding. There are no adequate safety data. To be precise about the reasoning: absence of data is not evidence of harm, and no regulatory contraindication was verified. The argument is about the shape of the ledger — an elective exposure with an unquantified allergy risk, no demonstrated benefit, and no safety data in these groups is not a trade that makes sense.

Proceed only with medical advice:
• Anyone on warfarin or another anticoagulant. One case report, causality graded possible, in a heavily co-medicated patient. That is thin evidence for a general interaction and sufficient reason to tell a prescriber.
• Anyone on glucose-lowering medication. Royal jelly has been flagged as potentially glucose-lowering at high doses. Note that the pooled trial evidence found no glycaemic effect at all, so this is a theoretical stacking caution, not a demonstrated benefit.
• Anyone on blood pressure medication. Flagged in a safety assessment; no case-level evidence located.

What is documented at case-report level and should be known but not inflated: haemorrhagic colitis (one case), contact dermatitis (listed in secondary reviews, no primary source verified here), and occupational asthma from inhaled royal jelly powder in processing and laboratory workers.

The honest statement about incidence, in both directions. The rate of severe reaction is not established. No incidence study exists. The Australia–New Zealand label says "in rare cases" and that is a regulator's phrasing, not a measurement, so this entry does not repeat it as a rate. What exists is a scattered case-report literature with no denominator, which is weak evidence that severe events are uncommon in people without asthma or atopy — weak because supplement adverse events are under-reported and because the cases cluster around one Australian allergy community in one decade. A separate figure does exist for sensitisation: about 7 percent skin-test positivity in a Hong Kong hospital-employee sample and in asthma-clinic attendees. Sensitisation is not clinical allergy and that figure is almost certainly many times higher than any real reaction rate; the study's own authors say the relationship remains to be defined.

And the counterweight that must ship with all of the above: in the clinical trials conducted, no adverse events were reported — none over twelve weeks at 800 mg per day in the menopause trial, none over three months of topical use in the foot ulcer trial. For a non-atopic adult at supplement doses, the short-term tolerability record in trials is genuinely clean. It is also, at twenty to a hundred participants per trial, structurally incapable of detecting the one serious risk that exists. Both of those sentences are true and neither should be dropped.

Controversy

Nature of the dispute. There are really two disputes stacked on top of each other. The narrow one is scientific and lives inside bee research: does a royal jelly protein actively make queens, or does queen fate depend on the absence of dietary inhibitors, or on chromatin-level effects, or on quantity? The broad one is commercial and lives in the supplement market: does any of the bee biology license any human claim? The narrow dispute is genuine and unresolved among researchers with no stake in supplement sales. The broad dispute is much less balanced than the marketing suggests.

Position A — royal jelly is a bioactive functional food whose potential is under-researched. The case: royal jelly demonstrably drives one of the most dramatic developmental switches in biology. It contains compounds found nowhere else in nature. Randomised placebo-controlled human trials exist and some are positive — anxiety and back pain in postmenopausal women, total antioxidant capacity up and malondialdehyde down across independent trials, a total-cholesterol reduction reported in two separate meta-analyses. Reviewers rate individual-trial risk of bias as mostly low. The reason the evidence base is thin is not that royal jelly failed adequate trials; it is that adequate trials have never been funded, because no patent protects the product. EFSA did not find royal jelly ineffective — it declined to assess the claims because the applicants could not characterise the substance well enough, which is a paperwork failure in a process where 421 of 442 claims failed. Meanwhile the digestion argument proves less than it appears: peptide fragments generated by digestion are themselves a legitimate research line.

Position B — royal jelly is insect biology sold as human transformation, with a real and under-disclosed harm. The case: the entire consumer proposition rests on an inference nobody has to state and nobody can therefore be held to. The mechanism that would bridge bee to human is contested in the primary literature — and the challenge was not merely a failed replication but a designed experiment in which depleting or enriching larval food with royalactin changed nothing. Its central precondition, intact protein reaching human tissue, was examined once and found unmet. The human trials test unrelated endpoints, are small enough that pooled samples run to a few hundred people, are geographically concentrated in producing regions, and in the one case with a hard clinical endpoint (diabetic foot ulcer healing) the result was negative. The most-cited positive trial was funded and largely written by a bee-products company. The inflammation meta-analysis reports heterogeneity near 98 percent, which makes its pooled estimates close to uninterpretable regardless of the certainty label attached. The lipid dose-response, in the one review that examined it, ran the wrong way on HDL and triglycerides. And uniquely among common supplements, there is a death in the literature and a government-mandated warning label — which almost no marketing mentions.

The funding and bias dimension — cui bono, both ways.

Running toward royal jelly: the trade is substantial and the product is premium-priced, and the flagship positive human trial for the strongest indication was funded by a royal jelly manufacturer with five author-employees. The funding was disclosed in the paper, to the authors' credit, and is almost never disclosed when the trial is cited. The entire human evidence base clusters in Iran and Japan, both major producing and consuming regions with established apitherapy traditions, plus one Slovenian contribution. That is not disqualifying — researchers study what is locally used, exactly as olive oil is studied in Spain — but it means the whole literature shares a cultural prior about the substance's value, which bears on outcome selection and on willingness to publish nulls. The queen-bee story itself is the most efficient marketing asset in the category: free, memorable, verifiable, and irrelevant to the buyer. Premium producers also benefit from a strict 10-HDA standard, and laboratories selling authentication benefit from adulteration being seen as widespread.

Running against royal jelly: this direction is unusually thin, which is worth saying plainly. Nobody sells "royalactin does nothing". The researchers who published the failed replication had an academic incentive — overturning a Nature paper is a career good — but no commercial one. The allergy researchers publishing cross-reactivity findings are in public hospitals. The regulator that mandates the warning label gets strong public-safety optics at near-zero enforcement cost, so mandated wording trends maximally cautionary, and that is a reason to quote it as a label rather than convert it into a rate. Competing supplement categories and generic scare content benefit from recycling the 1994 death, and a single fatality becomes disproportionately quotable. There is also a pharmaceutical-adjacent interest in framing all supplements as unregulated and dangerous, which the regulatory facts in this entry should not be read as endorsing.

The genuinely useful observation: the negative diabetic foot ulcer trial came out of Iran, a producing region. Producing-region origin has not uniformly suppressed null results, which is a real point in the literature's favour.

Realised Position: Royal jelly's most famous property is insect biology and it does not transfer, and the mechanism claimed to bridge it is disputed within bee research itself. The human trial literature is small, short, heterogeneous and mostly measured in blood markers rather than in anything a person would notice — with the single best-designed hard-endpoint trial returning a null. We do not conclude royal jelly is inert; we conclude nothing has been demonstrated, which is a different and more honest statement, and the digestion-product route remains a legitimate open question. What we do state without hedging is the shape of the evidence: for this supplement the human data is strongest on the harm side. There is a documented death in a person with asthma, a case literature concentrated in atopy, laboratory cross-reactivity with house dust mite and bee venom that can prime a first-dose reaction, and a government warning label — against a benefit case with no authorised health claim in the EU and no adequately powered trial anywhere. For a person with asthma or atopy the answer is no. For a non-atopic adult it is a low-probability risk with no demonstrated payoff, which is a bad trade rather than a dangerous one. We refuse both the transformation story and the blanket dismissal, and we consider the failure to state the asthma risk the more serious of the two common failures.

Cross-Pillar Connections

Diet. Royal jelly is the clearest case in the supplement aisle of a compound whose interest is real and whose dose is trivial: a one-gram capsule supplies roughly a tenth of a gram of protein against a daily intake measured in tens of grams. That arithmetic is the same argument made in whole_food_emphasis — the food matrix and the total intake do the work, and an isolated fraction at a fraction of a gram rarely changes anything. The 10-HDA question is a specific instance of the general form-and-dose reasoning in supplement_form_elemental_dose_and_bioavailability: a standardisation number on a label describes purity, not delivered dose at the tissue, and the two are routinely conflated.

The p-coumaric acid finding connects sideways to plant chemistry. If queen fate depends partly on the absence of plant phenolics from royal jelly, that is the mirror image of the argument in plant_defence_compounds_and_xenohormesis — where plant defence compounds are framed as mild stressors that provoke adaptive responses, here a plant phenolic acts as a developmental suppressor in an insect. Both entries are describing the same underlying fact: plants make compounds that alter animal physiology, and the sign of the effect depends entirely on the organism and the dose.

The antioxidant-marker literature is the same trap described elsewhere. Total antioxidant capacity up and malondialdehyde down, with no clinical outcome attached and heterogeneity near 98 percent, is precisely the pattern polyphenols_evidence_and_the_antioxidant_myth documents — markers move, outcomes do not follow, and the marker becomes the claim.

Immune and inflammation. Royal jelly is marketed for immune support and the EU claim submission included natural defence and immune system effects, which were not substantiated. The systems view in immune_function_cross_pillar_optimisation is the right frame for what actually moves immune function; a supplement whose CRP effect pooled to null does not appear in it. bovine_colostrum_gut_immune is the useful comparison case: another animal-derived secretion sold on the biology of its original recipient, with the same species-and-digestion questions and a different evidence profile.

Atopy is the pillar-crossing risk. People with hay fever, asthma or eczema are simultaneously the group most likely to be shopping for natural immune support and the group in which royal jelly's severe reactions concentrate. That overlap is not incidental, and it is why the safety section leads rather than trails. Management of the underlying conditions belongs in hay_fever_allergic_rhinitis_sleep_protocol and eczema_skin_barrier_lifestyle; the relevant instruction here is simply that royal jelly is contraindicated for those readers.

Menopause. The one indication with a positive human signal is also an indication with a far better-evidenced option. menopausal_hormone_therapy_evidence covers the comparison a person should actually be making; an 800 mg supplement with two significant endpoints out of two dozen in 42 women is not on the same evidential scale.

Testosterone. testosterone_optimization covers what does and does not move male hormonal status. Royal jelly's presence in that market rests on rodent damage-rescue models and on an unstated inference from insect caste biology, and it belongs in the "marketed, not evidenced" column.

What would change our mind

Industry bias note

Sources (51)

• *Composition and quality**↗
• Nutrient Composition and Quality Assessment of Royal Jelly Samples Relative to Feed Supplements. PMC11203284 (2024). Cited by title; author list not captured.↗
• ISO 12824:2016, Royal jelly — Specifications. Minimum 10-HDA thresholds; full standard text not read.↗
• 10-Hydroxy-2-Decenoic Acid (10-HDA) and Bioactive Components in Royal Jelly (review). Cited for 10-HDA range and uniqueness.↗
• *Caste biology and the mechanism dispute**↗
• Mao W, Schuler MA, Berenbaum MR. A dietary phytochemical alters caste-associated gene expression in honey bees. Science Advances (2015). doi:10.1126/sciadv.1500795↗, pubmed.ncbi.nlm.nih.gov/26601244↗/" target="_blank" rel="noopener">PMID 26601244↗.
• Kamakura M. Royalactin induces queen differentiation in honeybees. Nature 473:478–483 (2011). doi:10.1038/nature10093↗, pubmed.ncbi.nlm.nih.gov/21516106↗/" target="_blank" rel="noopener">PMID 21516106↗.
• Buttstedt A, Ihling CH, Pietzsch M, Moritz RFA. Royalactin is not a royal making of a queen. Nature 537(7621):E10–E12 (2016). doi:10.1038/nature19349↗, pubmed.ncbi.nlm.nih.gov/27652566↗/" target="_blank" rel="noopener">PMID 27652566↗. Brief Communications Arising; reports both failed replication attempts and a royalactin-depleted-versus-enriched feeding experiment, and extends the exclusion to MRJP2, 3 and 5.
• Kamakura M. Kamakura replies. Nature (2016). doi:10.1038/nature19350.↗ Existence confirmed; contents not read.
• PMC3059907 — histone deacetylase inhibitor activity in royal jelly and caste switching. Title-level verification only; unread; no author attributed.↗
• Communications Biology s42003-017-0004-4 — argues against a single master-inducer account. Title-level verification only; unread; no author attributed.↗
• PMC9506344 — queen versus worker comparison (lifespan, antioxidant, immune, gut microbiota). Cited by title.↗
• *Digestion and species transfer**↗
• Mureşan CI, Schierhorn A, Buttstedt A. The Fate of Major Royal Jelly Proteins during Proteolytic Digestion in the Human Gastrointestinal Tract. J Agric Food Chem 66(16):4164–4170 (2018). doi:10.1021/acs.jafc.8b00961↗, pubmed.ncbi.nlm.nih.gov/29629561↗/" target="_blank" rel="noopener">PMID 29629561↗. In vitro simulated digestion, not a human trial.
• pubmed.ncbi.nlm.nih.gov/11834223↗/" target="_blank" rel="noopener">PMID 11834223↗ — gastrointestinal enzyme treatment of royal jelly protein generating blood-pressure-active peptides in rats. Title-level verification only; rodent; effect sizes not stated here.
• *Human trials and reviews**↗
• Asama T, Matsuzaki H, Fukushima S, Tatefuji T, Hashimoto K, Takeda T. Royal Jelly Supplementation Improves Menopausal Symptoms Such as Backache, Low Back Pain, and Anxiety in Postmenopausal Japanese Women. Evid Based Complement Alternat Med 2018;2018:4868412. PMC5949161. Funded by Yamada Bee Company; five author-employees; one further author received a company advisory fee.↗
• Mahboobi S, Jafarnejad S, Eftekhari MH. Royal jelly does not improve markers of glycemia: A systematic review and meta-analysis of Randomized Clinical Trials. Complement Ther Med 2019;44:235–241. pubmed.ncbi.nlm.nih.gov/31126561↗/" target="_blank" rel="noopener">PMID 31126561↗.
• Siavash M, Shokri S, Haghighi S, Shahtalebi MA, Farajzadehgan Z. The efficacy of topical royal jelly on healing of diabetic foot ulcers: a double-blind placebo-controlled clinical trial. Int Wound J. pubmed.ncbi.nlm.nih.gov/23566071↗/" target="_blank" rel="noopener">PMID 23566071↗. Retrieved at abstract/listing level; author list to be re-checked before onward quotation.
• Taheri et al. Effects of royal jelly consumption on inflammation and oxidative stress: A systematic review and meta-analysis of randomized controlled trials. PMC12244952 (2025). 6 RCTs, 7 arms, 356 participants (181 intervention / 175 control). Abstract-level retrieval; only the first author was captured.↗
• Royal jelly is an effective and relatively safe alternative approach to blood lipid modulation: A meta-analysis. J Funct Foods (2018). Full text not retrieved; author list deliberately not asserted.↗
• The effects of Royal Jelly consumption on lipid profile: A GRADE-assessed systematic review and dose-response meta-analysis. Clin Nutr ESPEN (2023). Full text not retrieved; author list deliberately not asserted. Abstract reports total cholesterol WMD −8.75 mg/dL (95% CI −15.50 to −1.99, P=0.01, I²=68.5 percent), no effect on triglycerides, LDL or HDL overall, and a dose-response showing greater HDL reduction and greater triglyceride elevation at higher doses.↗
• PMC10438987 — royal jelly and anthropometric indices, GRADE-assessed dose-response meta-analysis (10 studies, 512 participants). Cited by title.↗
• Effect of Processed Honey and Royal Jelly on Cancer-Related Fatigue: A Double-Blind Randomized Clinical Trial. pubmed.ncbi.nlm.nih.gov/27504161↗/" target="_blank" rel="noopener">PMID 27504161↗ (n=52). Full text not retrieved; authors not attributed.
• Oral Intake of Royal Jelly Has Protective Effects Against Tyrosine Kinase Inhibitor-Induced Toxicity in Patients with Renal Cell Carcinoma. Medicines 2019;6(1):2 (n=33). Full text not retrieved; authors not attributed.↗
• Alzheimer's Drug Discovery Foundation, Cognitive Vitality assessment: Royal Jelly. Single human trial identified is a royal jelly + ginkgo + ginseng combination, n=66, 4 weeks.↗
• pubmed.ncbi.nlm.nih.gov/25050300↗/" target="_blank" rel="noopener">PMID 25050300↗ — royal jelly, sperm parameters, testosterone and lipid peroxidation in mice treated with oxymetholone.
• *Allergy, anaphylaxis and harm**↗
• Bullock RJ, Rohan A, Straatmans JA. Fatal royal jelly-induced asthma. Med J Aust 1994;160(1):44. doi:10.5694/j.1326-5377.1994.tb138207.x↗, pubmed.ncbi.nlm.nih.gov/8271989↗/" target="_blank" rel="noopener">PMID 8271989↗. Full text not read; no abstract is indexed. Patient described only as a person with pre-existing asthma in Australia — secondary sources conflict on age and sex.
• Thien FC, Leung R, Plomley R, Weiner J, Czarny D. Royal jelly-induced asthma. Med J Aust 1993;159(9):639. doi:10.5694/j.1326-5377.1993.tb138070.x↗, pubmed.ncbi.nlm.nih.gov/8123114↗/" target="_blank" rel="noopener">PMID 8123114↗. Listing-level verification.
• Thien FC et al. Asthma and anaphylaxis induced by royal jelly. Clin Exp Allergy 1996;26(2). pubmed.ncbi.nlm.nih.gov/8835130↗/" target="_blank" rel="noopener">PMID 8835130↗. Abstract not read; no subject count stated here.
• Leung R, Thien FC, Baldo B, Czarny D. Royal jelly-induced asthma and anaphylaxis: clinical characteristics and immunologic correlations. J Allergy Clin Immunol (1995). Full text inaccessible during research; NO finding in this entry is attributed to it.↗
• Leung R, Ho A, Chan J, Choy D, Lai CK. Royal jelly consumption and hypersensitivity in the community. Clin Exp Allergy 1997;27(3):333–336. pubmed.ncbi.nlm.nih.gov/9088660↗/" target="_blank" rel="noopener">PMID 9088660↗. All figures in this entry verified against the abstract. Note that the abstract characterises the adverse-reaction/clinical-allergy association as positive although its confidence interval crosses one; this entry reports the interval.
• Hata T, Furusawa-Horie T, Arai Y, Takahashi T, Seishima M, Ichihara K. Studies of royal jelly and associated cross-reactive allergens in atopic dermatitis patients. PLOS ONE 2020;15(6):e0233707. doi:10.1371/journal.pone.0233707.↗ States that honeybee venom contains MRJP8 and MRJP9, registered as Api m 11.
• pubmed.ncbi.nlm.nih.gov/35221696↗/" target="_blank" rel="noopener">PMID 35221696↗ — anaphylaxis caused by major royal jelly protein 3. Cited for existence.
• pubmed.ncbi.nlm.nih.gov/9257239↗/" target="_blank" rel="noopener">PMID 9257239↗ — haemorrhagic colitis associated with royal jelly intake. Author list not verified; no authors named.
• Lee NJ, Fermo JD. Warfarin and royal jelly interaction. Pharmacotherapy 2006;26(4):583. doi:10.1592/phco.26.4.583↗, pubmed.ncbi.nlm.nih.gov/16553520↗/" target="_blank" rel="noopener">PMID 16553520↗. Causality graded possible.
• Fantini P, Delle Donne P, Calogiuri G, et al. Oral Allergy Syndrome in a Child Provoked by Royal Jelly. Case Rep Med 2014;2014:941248. pubmed.ncbi.nlm.nih.gov/24799914↗/" target="_blank" rel="noopener">PMID 24799914↗, PMC3988729. Seven-year-old child; sex not specified in the record and not asserted here.
• Occupational royal jelly asthma reports (respiratory disease in a processing worker; laboratory workers with inhalation-challenge confirmation). Titles and route confirmed; full citations not retrieved; one item is a conference abstract. No fabricated citation given.↗
• *Regulatory**↗
• Food Standards Australia New Zealand, Warning and advisory statements. Royal jelly warning quoted verbatim; applies in both Australia and New Zealand and may not be paraphrased on-label. No Food Standards Code clause number cited — none confirmed.↗
• EFSA Panel on Dietetic Products, Nutrition and Allergies. Scientific Opinion on health claims related to various foods/food constituents including royal jelly (IDs 1225, 1226, 1227, 1228, 1230, 1231, 1326, 1328, 1329, 1982, 4696, 4697), Article 13(1) of Regulation (EC) No 1924/2006. EFSA Journal 2011;9(4):2083. doi:10.2903/j.efsa.2011.2083.↗ Full text paywalled during research; no verbatim conclusion quoted.
• Therapeutic Goods Administration (Australia), royal jelly advertising permission. Instrument exists; wording not retrieved.↗
• Directive 2002/46/EC (EU food supplements); Dietary Supplement Health and Education Act 1994 (US); UK Food Standards Agency regulated products guidance. Assembled from regulatory summaries, not primary instruments.↗
• *Explicitly not cited, and why**↗
• Any incidence or prevalence figure for severe reaction — none was located anywhere.↗
• The "roughly one in six adults attending asthma clinics have royal-jelly-reactive IgE" figure — stated generically in the text, could not be tied to a named study.↗
• A commercial-sample 10-HDA failure rate — the likely source was inaccessible; no number given.↗
• Human testosterone or fertility trial figures — search summaries returned mutually incompatible doses for what may be the same study.↗
• The sex of the child in the Fantini paediatric case, and the age or sex of the patient in the fatal case — neither is confirmable from the retrievable record.↗
• A 72-week zebrafish study on a branded royal jelly product, and a murine model in which royal jelly prevented anaphylactic response. Both were encountered during research and are named here only so they are never cited toward a human claim; the murine result points in the opposite direction from all human allergy evidence and must not appear near a human safety statement.↗