The Etalyn compound register

Modafinil

One dose makes healthy, well-rested adults think more sharply than a dummy pill does.

Claim as graded “Acute modafinil improves overall cognitive performance in healthy, non-sleep-deprived adults versus placebo”

Evidence tier A · provenance grade P3 · direction Supported · retail band 1, Strong evidence. Derived by rule 6 - tier A/B with provenance P3+. Group: Pharmaceuticals, nootropics and controlled substances. Scope qualifier: statistically significant but trivially small; attenuates further after publication-bias correction

Best available evidence: Kredlow et al. meta-analysis of placebo-controlled trials of modafinil as a cognitive enhancer; corroborated by Roberts et al. 2020 meta-analysis in Eur Neuropsychopharmacol

Sample and design: 19 placebo-controlled trials, 26 subgroups, 406 comparisons → 67 domain-specific effect sizes, 767 participants (Kredlow et al.); independently, 14 studies / 64 effect sizes, 260 participants in repeated-measures designs and 312 in between-groups designs (135 modafinil, 177 placebo) (Roberts et al. 2020)

Effect as reported: Overall Hedges' g = 0.10 (95% CI 0.05 to 0.15), P < 0.001, n = 67 effect sizes, I² = 0%; Trim-and-Fill adjusted g = 0.07 (0.02 to 0.12); by domain: processing speed g = 0.20 (0.07 to 0.33) P < 0.01, executive functioning g = 0.10 (0.01 to 0.18) P < 0.05, memory g = 0.07 (−0.02 to 0.16) P = 0.14, attention g = 0.06 (−0.06 to 0.18) P = 0.32 (Kredlow et al.). Independently SMD = 0.12 (95% CI 0.02 to 0.21), Z = 2.45, p = 0.01, I² = 72%, with only the "updating" domain significant (SMD 0.28, 0.02 to 0.54, p = 0.03) and sustained attention null (SMD −0.13, −0.52 to 0.26, p = 0.52) (Roberts et al. 2020)

Regulatory position: Licensed indication is not cognitive enhancement. FDA-approved PROVIGIL label: "indicated to improve wakefulness in patients with excessive sleepiness associated with narcolepsy, obstructive sleep apnea/hypopnea syndrome, and shift work sleep disorder"; "Rx Only"; "Modafinil (PROVIGIL) is listed in Schedule IV of the Controlled Substances Act" (FDA label). In the EU the CHMP Article 31 referral narrowed this further: "modafinil is only indicated to treat narcolepsy", removing obstructive sleep apnoea, shift-work sleep disorder and idiopathic hypersomnia, confirmed on re-examination 18 November 2010 (EMA referral)

Safety signal: Serious cutaneous reactions: 16 post-marketing cases of Stevens-Johnson syndrome, toxic epidermal necrolysis or erythema multiforme, three of them fatal, causality not excludable for the majority; three further serious cutaneous adverse reactions in clinical trials, all in children; rash leading to discontinuation in 13/1,585 paediatric patients (~0.8%) versus 0/4,264 adults; "Modafinil should be discontinued at the first sign of rash and should not be restarted" (EMA Article 31 annexes). EMA advises against use in uncontrolled hypertension or irregular heart beat and against use in children (EMA referral)

Sources: Kredlow meta-analysis Roberts et al. 2020 full text FDA PROVIGIL label EMA modafinil referral EMA referral annexes

Methylphenidate

One dose makes healthy adults think more sharply than a dummy pill does.

Claim as graded “Acute methylphenidate improves overall cognitive performance in healthy adults versus placebo”

Evidence tier B · provenance grade P3 · direction Supported · retail band 1, Strong evidence. Derived by rule 6 - tier A/B with provenance P3+. Group: Pharmaceuticals, nootropics and controlled substances. Scope qualifier: small; domain-specific, largely confined to recall

Best available evidence: Roberts et al. 2020 meta-analysis, Eur Neuropsychopharmacol

Sample and design: 24 studies, 47 effect sizes; "501 participants from repeated measures designs and 144 from between-groups designs (92 MPH, 92 placebo)" (Roberts et al. 2020

Effect as reported: Overall SMD = 0.21 (95% CI 0.09 to 0.32), Z = 3.54, p = .0004, I² = 66%; driven by recall SMD = 0.43 (0.20 to 0.65), Z = 3.70, p = 0.0002, I² = 0% and inhibitory control SMD = 0.27 (0.02 to 0.51), p = 0.03; null for updating 0.06 (−0.24 to 0.37) p = 0.67, switching 0.02 (−0.14 to 0.18) p = 0.80, spatial working memory −0.14 (−0.50 to 0.21) p = 0.43, selective attention 0.03 (−0.36 to 0.42) p = 0.88; subgroup difference χ² = 18.27, df = 7, p = .01; Egger's t(46) = 1.62, p = 0.11 (Roberts et al. 2020). A separate double-blind trial in 36 healthy men aged 18–30 given single oral doses of 10, 20 or 40 mg or placebo (n = 9 per arm) reported "No differences in performance were observed on any of the tests" (Batistela et al.)

Regulatory position: Licensed indication is not cognitive enhancement in healthy adults. FDA Ritalin label: indicated for "Attention Deficit Disorders, Narcolepsy" and "as an integral part of a total treatment program"; product type "HUMAN PRESCRIPTION DRUG"; DEA Schedule "CII"; "RITALIN® is a federally controlled substance (CII) because it can be abused or lead to dependence"; "Selling or giving away RITALIN® may harm others, and is against the law" (DailyMed Ritalin label)

Safety signal: Federally controlled Schedule II stimulant carrying an explicit abuse-and-dependence statement on its FDA-approved Medication Guide (DailyMed Ritalin label). Quantified cardiovascular or psychiatric adverse-event rates in healthy off-label users: n.a.

Sources: Roberts et al. 2020 full text Batistela et al., PMC5642404 DailyMed Ritalin label

Clenbuterol

It strips fat off healthy adult men.

Claim as graded “Clenbuterol reduces fat mass in healthy adult men”

Evidence tier D · provenance grade P4 · direction Not supported · retail band 5, Evidence says no. Derived by rule 1 - direction Not supported. Group: Pharmaceuticals, nootropics and controlled substances. Scope qualifier: for fat loss; lean-mass accretion was observed in the same trial

Best available evidence: Hostrup group randomised, double-blinded, placebo-controlled cross-over trial, University of Copenhagen, Nov 2021–Dec 2022, registered NCT03860870

Sample and design: 11 healthy men aged 18–40 completed (13 eligible, 2 withdrew, 1 partial); double-blinded placebo-controlled cross-over; two 2-week cycles with 3-week washout; 80 µg/day was administered (4 × 20 µg, Spiropent); DXA body composition (PMC12487599)

Effect as reported: Fat mass 0.00 kg (95% CI −0.52 to 0.52), P = 0.994; body fat percentage −0.20% (−0.79 to 0.39), P = 0.501; by contrast lean mass +0.91 kg (95% CI 0.02 to 1.81), P = 0.046, and aerobic capacity fell: absolute V̇O₂max −249 mL/min (−378 to −120), P < 0.001, relative V̇O₂max −3.6 mL/min/kg (−5.4 to −1.9) P < 0.001, Wmax −16 W (−23 to −8) P < 0.001 — a 7% reduction in maximal oxygen uptake and 4% in exercise capacity (PMC12487599)

Regulatory position: "the Food and Drug Administration (FDA) has never approved clenbuterol for human use under the Federal Food, Drug and Cosmetic Act (FD&C Act) due to unacceptable safety risks that include significant adverse cardiovascular and neurological effects"; the only US approval is veterinary — Ventipulmin® Syrup, approved 1998 for airway obstruction in horses; "Outside the United States, clenbuterol may be prescribed for the treatment of bronchial asthma in humans"; "Clenbuterol is currently not controlled under the Controlled Substances Act (CSA)" (DEA Diversion Control). In the EU, Council Directive 96/22/EC "prohibits the use of clenbuterol in all farm animals with the exception of some specific therapeutic purposes in equines and in cows" (Commission Regulation (EC) No 1312/96); the human medicine Spiropent (Hikma, Portugal) was the product administered in the cited trials (Nat Commun 2023)

Safety signal: FDA position quoted above cites "unacceptable safety risks that include significant adverse cardiovascular and neurological effects" (DEA). The trial itself found a statistically significant loss of aerobic capacity (V̇O₂max −249 mL/min, P < 0.001) (PMC12487599). A fatality case report titled "Death of an apprentice bodybuilder following 2,4-dinitrophenol and clenbuterol intake" is indexed in the DNP toxicology literature (PMC12756549)

Sources: PMC12487599 (NCT03860870 results) Nat Commun 2023 (NCT03800290) ClinicalTrials.gov NCT03860870 DEA clenbuterol factsheet Commission Reg (EC) 1312/96

Liothyronine (T3)

It causes weight loss in adults whose thyroid is working normally.

Claim as graded “Liothyronine produces weight loss in euthyroid adults”

Evidence tier D · provenance grade P3 · direction Insufficient · retail band 6, Too little evidence yet. Derived by rule 3 - direction Insufficient. Group: Pharmaceuticals, nootropics and controlled substances.

Best available evidence: Kaptein EM, Beale E, Chan LS., J Clin Endocrinol Metab 2009;94(10):3663–3675 — systematic review of thyroid hormone therapy in euthyroid patients, structured abstract on NCBI Bookshelf (DARE)

Sample and design: "Twenty-one studies (n=491) were included in the review including 13 RCTs"; 14 of these addressed obesity (6 RCTs, 8 observational); T3 was administered at 25–28 µg/70 kg/day for 7–21 days (DARE structured abstract, NBK77205)

Effect as reported: No numeric effect reported. The review states: "Consistent effects of T3 and T4 therapy on weight loss, leucine and ketone metabolism, oxygen consumption and cardiac outcomes were not established" (NBK77205). A supporting double-blind RCT in which T3 60 µg/day was administered alongside a 320 kcal/day formula diet for 12 weeks reported that "T3-treated patients had a significantly greater weight-loss by the 12th week" but stated no participant count, effect size, CI or p-value (Moore et al., Lancet 1980, PMID 6101677)

Regulatory position: Licensed indication is not weight loss. UK SmPC: "Liothyronine is indicated in adults and children for the treatment of coma of myxedema, the management of severe chronic thyroid deficiency and hypothyroid states occurring in the treatment of thyrotoxicosis"; also "as an adjunct to carbimazole to prevent sub-clinical hypothyroidism developing during carbimazole treatment of thyrotoxicosis" (emc SmPC, Liothyronine Sodium 20 microgram Tablets). Controlled-substance scheduling: not stated on the SmPC; not checked elsewhere

Safety signal: The review authors wrote that "thyroid hormone therapy should be discouraged in euthyroid patients given the insufficient data available"; it also found "some evidence that thyroid hormone therapy can result in subclinical hypothyroidism", and that "Mortality was assessed in four studies and in one study was significantly increased 3.3-fold with T4 administered at 300μg/70kg per day for two days in patients with acute renal failure" (NBK77205)

Sources: DARE structured abstract of Kaptein et al. 2009 (NBK77205) Moore et al., Lancet 1980, PMID 6101677 emc liothyronine SmPC

2,4-Dinitrophenol (DNP)

It causes weight loss in adults with obesity.

Claim as graded “DNP produces weight loss in obese adults”

Evidence tier D · provenance grade P2 · direction Supported · retail band 6, Too little evidence yet. Derived by rule 4 - tier D or E. Group: Pharmaceuticals, nootropics and controlled substances. Scope qualifier: weight loss is real and dose-proportional; it is produced by uncontrolled mitochondrial uncoupling and is inseparable from the lethality below

Best available evidence: ATSDR toxicological profile summary of the 1930s clinical literature, plus the modern mortality record from Grundlingh et al. (J Med Toxicol), Kamour et al. (Emerg Med J 2015) and the UK National Poisons Information Service

Sample and design: Historical, uncontrolled: 37 obese patients administered 1 mg/kg/day as the sodium salt for an average of 14 days; 159 patients administered ~3 mg/kg/day for 22–89 days; 170 patients of whom 100 took at least 4 mg/kg/day for at least 6 weeks (average 88 days); 4 volunteers administered ~4 mg/kg/day for 7–16 days — none placebo-controlled (ATSDR profile, NBK590059). Mortality: 62 published deaths reviewed (Grundlingh et al., PMC3550200); 30 systemic exposures 2007–2013 to the UK NPIS (Kamour et al., PMID 24957806); 148 NPIS cases since 2007 (NPIS Report 2022–23)

Effect as reported: Average weight loss of 0.43 kg/week in the 37 obese patients; "Average weight losses of approximately 0.5–1 kg/week" across obese or psychiatric patients administered 2–4 mg/kg/day for 1 week to 18 months; average basal metabolic rate increased by approximately 11% for each 100 mg (1 mg/kg) increase in dose in the 159-patient series (ATSDR NBK590059). No confidence intervals or p-values exist in this literature. The modern review likewise states "Weight loss of up to 1.5 kg per week is reported" and "An average metabolic rate increase of 11% for every 100 mg of DNP when taken regularly" (PMC3550200)

Regulatory position: "The FDA has never approved 2,4-DNP as a pharmaceutical agent." "In 1938, the FDA declared DNP to be 'extremely dangerous' and 'not fit for human consumption'" (ATSDR NBK590059). UK: DNP was added to the regulated poisons list under the Control of Explosives Precursors and Poisons 2023; as of 1 October 2023 a member of the public must hold a Home Office Explosives Precursors and Poisons (EPP) licence to import, acquire, possess or use it, and "It is now a criminal offence to sell this substance to members of the public without a valid Home Office issued EPP licence" (UK Food Standards Agency)

Safety signal: Extreme — fatal. "There were five (17%, 95% CI 6.9% to 34%) fatalities" among 30 UK systemic exposures 2007–2013 (Kamour et al., PMID 24957806). "In total there have now been 148 cases of systemic DNP exposure discussed by phone with the NPIS since 2007… Of these, 26 (17.6%) are known to have died"; "there have been at least 33 DNP-related deaths in the UK since 2007, including 26 since January 2015"; "These effects can be fatal in spite of intensive medical treatment" (NPIS Report 2022–23). 62 published deaths, including 36 in one 1919 Paris munitions-factory report and 12 fatalities in 2001–2010; the lowest published lethal human oral dose was 4.3 mg/kg; doses in published acute/suicidal fatalities ranged 2.8 g to an estimated 5 g; average time to death 14 hours; "No patient has been recorded to be asymptomatic beyond 10 h after an acute overdose" (Grundlingh et al., PMC3550200). A recent Swedish fatal case died 4 hours after ingesting 5 g, with core temperature 40 °C, serum potassium 11 mmol/L and peri-mortem rigidity "consistent with catastrophic ATP depletion"; "nine forensically confirmed DNP-related deaths have occurred since 2010" in Sweden (PMC12756549). FSA: "DNP is poisonous to humans and can cause death"; "Taking DNP has resulted in a significant number of deaths in the UK" (FSA)

Sources: ATSDR toxicological profile, NBK590059 Grundlingh et al., PMC3550200 Kamour et al., Emerg Med J 2015, PMID 24957806 NPIS Report 2022–23 UK Food Standards Agency PMC12756549 (Swedish cases)

Ephedrine / ephedra alkaloids

Products containing it cause weight loss in adults who are overweight or have obesity.

Claim as graded “Ephedrine-containing products produce weight loss in overweight or obese adults”

Evidence tier B · provenance grade P3 · direction Supported · retail band 1, Strong evidence. Derived by rule 6 - tier A/B with provenance P3+. Group: Pharmaceuticals, nootropics and controlled substances. Scope qualifier: short-term only; ~2 kg, with no data beyond 6 months

Best available evidence: Meta-analysis of ephedrine-containing products for weight loss (PMC8618781); corroborated by the RAND/Shekelle evidence report as summarised by the NIH Office of Dietary Supplements

Sample and design: "Nine studies with 534 patients were evaluated to investigate the weight loss effect of ephedrine-containing products in obese or overweight individuals" (PMC8618781); separately 44 controlled trials identified with 20 entering the RAND meta-analysis (NIH ODS)

Effect as reported: MD = −1.97 kg (95% CI −2.38 to −1.57; p < 0.00001; I² = 80%); post-2000 subgroup of 5 studies MD = −2.25 kg (−2.81 to −1.69; p < 0.00001; I² = 88%); heart rate increased by +5.76 beats/min (95% CI 3.42 to 8.10), p < 0.00001; Begg's p = 0.677, Egger's p = 0.356 (PMC8618781). RAND: ephedrine promoted weight loss of about 1.3 pounds/month more than placebo (5 studies), ephedrine plus caffeine 2.2 pounds/month more (12 studies), ephedra plus caffeine-containing herbs 2.1 pounds/month more (4 studies); "No studies assessed the long-term effects on weight loss; the longest published follow-up was 6 months" (NIH ODS)

Regulatory position: "On February 11, 2004, FDA published in the Federal Register a final rule that established a regulation declaring dietary supplements containing ephedrine alkaloids adulterated under the Federal Food, Drug, and Cosmetic Act because they present an unreasonable risk of illness or injury… (69 FR 6787)"; "The final rule became effective on April 12, 2004"; "It is illegal to market an adulterated dietary supplement" (FDA small entity compliance guide, docket FDA-1995-N-0054); codified at 21 CFR § 119.1, citation "[69 FR 6853, Feb. 11, 2004]". The meta-analysis notes that "the U.S. District Court overturned the FDA's ban" one year later (PMC8618781). EU: "Ephedra herb and its preparations originating from Ephedra species" was added to Part A of Annex III of Regulation (EC) No 1925/2006 — the list of prohibited substances — by Commission Regulation (EU) 2015/403 of 11 March 2015

Safety signal: FDA acted "based upon the well-known pharmacology of ephedrine alkaloids, the peer-reviewed scientific literature… and the adverse events reported… (69 FR at 6788)" and found an "unreasonable risk of illness or injury" (FDA). Adverse events were meta-analysed from 50 of 52 RCTs, showing a 2–3 fold increase in nausea, vomiting, psychiatric symptoms such as anxiety and mood change, autonomic hyperactivity and palpitations (NIH ODS). Measured heart-rate increase of +5.76 bpm (3.42 to 8.10), p < 0.00001 (PMC8618781)

Sources: PMC8618781 meta-analysis NIH ODS ephedra/ephedrine fact sheet FDA final-rule compliance guide 21 CFR § 119.1 Commission Reg (EU) 2015/403

Yohimbine (and rauwolscine)

It adds extra weight or fat loss for adults already eating fewer calories.

Claim as graded “Yohimbine increases weight or fat loss in adults on an energy-restricted diet”

Evidence tier D · provenance grade P3 · direction Contested · retail band 4, Experts disagree. Derived by rule 2 - direction Contested. Group: Pharmaceuticals, nootropics and controlled substances. Scope qualifier: one small positive double-blind trial; a same-era trial titled for lack of efficacy; no pooled analysis found

Best available evidence: Kucio C, Jonderko K, Piskorska D., "Does yohimbine act as a slimming drug?", Isr J Med Sci 1991;27(10):550–6

Sample and design: "Twenty female obese outpatients were subjected to a 3-week low-energy diet (1,000 kcal/day), after which they were randomly allocated according to a double-blind study protocol to two treatments: 10 subjects received 5 mg yohimbine per os 4 times a day and 10 received a placebo for 3 weeks, in addition to a low-energy diet of 1,000 kcal/day" (PubMed 1955308)

Effect as reported: "Yohimbine significantly increased the mean weight loss in patients on a low-energy diet: 3.55 +/- 0.24 kg (yohimbine) vs. 2.21 +/- 0.37 kg (placebo), P less than 0.005"; no confidence interval was reported; "No significant effect of yohimbine on lipolysis was observed under the experimental conditions of this study" (PubMed 1955308). A separate trial titled "[Lack of efficacy of yohimbine in the treatment of obesity]" randomised participants to yohimbine (n = 10, 18 mg/day) or placebo (n = 9) (PubMed 3795978 — abstract text could not be retrieved this session)

Regulatory position: No FDA- or EMA-approved indication for weight loss was found. EU: "Yohimbe bark and its preparations originating from Yohimbe (Pausinystalia yohimbe (K. Schum) Pierre ex Beille)" was added to Part C of Annex III of Regulation (EC) No 1925/2006 — substances under Union scrutiny — because EFSA "concluded that the chemical and toxicological characterisation of yohimbe bark and its preparations used in food… are not adequate to conclude on their safety as ingredients of food" and "it was not possible for the Authority to provide advice on a daily intake of yohimbe bark and its preparations that does not give rise to concerns for human health" (Commission Regulation (EU) 2015/403). Rauwolscine regulatory status: n.a.

Safety signal: "Data from poison control and food safety centers internationally have shown a notable increase in reported adverse events associated with YHM-containing products over the past two decades"; "YHM has a well-documented dose-dependent risk profile, with higher doses linked to gastrointestinal distress, hypertension, tachycardia, anxiety, sweating, and headaches"; "In rare cases, severe reactions such as seizures, loss of consciousness, and even death have been reported"; adverse events "have occurred when YHM is consumed in large quantities or combined with other stimulants (e.g., caffeine, synephrine, rauwolscine), exacerbating its sympathomimetic actions" (PMC11677475). Numeric poison-centre case and fatality counts: n.a.

Sources: Kucio et al. 1991, PubMed 1955308 PubMed 3795978 Ergogenic and Sympathomimetic Effects of Yohimbine, PMC11677475 Commission Reg (EU) 2015/403

Nicotinamide mononucleotide (NMN)

Taking it lowers fasting blood sugar in adults.

Claim as graded “NMN supplementation improves cardiometabolic markers (fasting blood glucose) in adults”

Evidence tier B · provenance grade P3 · direction Not supported · retail band 5, Evidence says no. Derived by rule 1 - direction Not supported. Group: Pharmaceuticals, nootropics and controlled substances.

Best available evidence: Meta-analysis of randomised controlled trials of NMN supplementation on glycaemic and lipid markers, Crit Rev Food Sci Nutr (PROSPERO CRD42023482534)

Sample and design: 12 RCTs, 513 participants (304 NMN, 209 placebo); 250–1250 mg/day administered for 2–24 weeks; mean ages 35–81; risk of bias (RoB2): 7 studies some concerns, 5 studies high risk (Crit Rev Food Sci Nutr)

Effect as reported: Fasting blood glucose MD = −0.39 mg/dL (95% CI −2.52 to 1.75), p = 0.683, I² = 0% (7 studies); all lipid outcomes null — triglycerides −7.26 mg/dL (−20.40 to 5.89) p = 0.250; total cholesterol −4.01 (−14.35 to 6.33) p = 0.397; LDL-C −2.11 (−9.09 to 4.87) p = 0.515; HDL-C −1.30 (−3.92 to 1.31) p = 0.293 (Crit Rev Food Sci Nutr). A parallel systematic review of 10 studies / 437 patients (mean age 58.0 years, mean follow-up 9.6 weeks) reported that "A meta-analysis was intended to be performed; however, this was not possible due to the significant heterogeneity of the included studies", that participants "demonstrated non-significantly improved physical performance parameters", and that "body composition and muscle mass are not significantly affected" (PMC11365583)

Regulatory position: "FDA initiated a review of past notification responses for NMN and concluded that NMN is excluded from the definition of a dietary supplement. This means that NMN may not be marketed as or in a dietary supplement." The stated basis is 21 U.S.C. § 321(ff)(3)(B)(ii): "NMN is an article authorized for investigation as a new drug by the FDA" (FDA letter to SyncoZymes re NDIN 1240 and 1247, 4 November 2022). No FDA or EMA marketing authorisation for NMN as a medicine was found; controlled-substance status: none found

Safety signal: The systematic review reported NMN "is well tolerated with no serious adverse effects" across 10 studies / 437 patients (PMC11365583). The principal regulatory signal is legal rather than toxicological: FDA has stated NMN "may not be marketed as or in a dietary supplement" (FDA). Long-term (>24 week) safety data: n.a.

Sources: Crit Rev Food Sci Nutr meta-analysis Systematic review, PMC11365583 FDA NMN NDI response letter

Rapamycin (sirolimus)

Taken occasionally at a low dose, it reduces belly fat around the organs in healthy adults aged 50 to 85.

Claim as graded “Low-dose intermittent rapamycin improves healthspan markers — primary endpoint visceral adipose tissue — in healthy adults aged 50–85”

Evidence tier D · provenance grade P4 · direction Not supported · retail band 5, Evidence says no. Derived by rule 1 - direction Not supported. Group: Pharmaceuticals, nootropics and controlled substances. Scope qualifier: primary endpoint; one secondary lean-mass signal in women only

Best available evidence: PEARL trial: "a decentralized, single-center, prospective, double-blind, placebo-controlled trial", NCT04488601, published in Aging-US

Sample and design: 114 participants completed and were analysed (10 mg n = 35; 5 mg n = 40; placebo n = 39), 11 discontinued; healthy adults aged 50–85 or with well-managed stable chronic disease; 48 weeks; compounded rapamycin administered orally once weekly; DXA body composition (Aging-US article 206235)

Effect as reported: Primary endpoint failed: "No significant differences were found for the primary endpoint of VAT after 48 weeks for either gender." Odds of VAT improvement were null in all participants (10 mg OR 1.68, 95% CI 0.66 to 4.32, p = 0.28; 5 mg OR 1.81, 0.73 to 4.50, p = 0.20). The only positive body-composition finding was a secondary endpoint in women: lean tissue mass at 48 weeks, 10 mg versus placebo md = 6.19390 (95% CI 0.8773 to 11.5105), p = 0.018, and 10 mg versus 5 mg md = 5.56454 (0.5311 to 10.5979), p = 0.026; male lean tissue mass was null at both timepoints (48 w F(2, 54) = 0.379, p = 0.686). Bone mineral density moved the wrong way for the 5 mg group across all participants, OR 0.24 (0.06 to 0.93), p = 0.04. A bioavailability caveat applies: the compounded rapamycin used had "approximately ⅓ the concentration in blood after 24 hrs relative to commercial" (Aging-US article 206235)

Regulatory position: Licensed indication is not longevity or healthspan. Rapamune (sirolimus) holds an EU marketing authorisation valid throughout the EU since 14 March 2001, EMEA/H/C/000273, indicated for prophylaxis of organ rejection in adult renal transplant patients and for sporadic lymphangioleiomyomatosis; it is subject to medical prescription (EMA Rapamune EPAR)

Safety signal: Within PEARL: serious adverse events numbered 1 (10 mg), 2 (5 mg) and 3 (placebo); non-severe AEs were similar across groups (10 mg = 117, 5 mg = 116, placebo = 122) though "GI symptoms were reported more often for rapamycin users than placebo (10 mg = 8, 5 mg = 7, placebo = 4)"; there was "a single report of anemia in the entire study, in a participant in the 5 mg treatment group", resolved with blood transfusion. The bone-density signal — 5 mg OR 0.24 (0.06 to 0.93), p = 0.04 — is an adverse direction (Aging-US article 206235). Immunosuppression-related risks from the licensed transplant setting are not quantified here: n.a.

Sources: PEARL trial, Aging-US article 206235 medRxiv preprint EMA Rapamune EPAR

Psilocybin (low-dose / "microdosing")

Tiny doses lift mood, wellbeing or thinking in healthy adults, compared with a dummy pill.

Claim as graded “Psilocybin microdosing improves mood, wellbeing or cognition in healthy adults relative to placebo”

Evidence tier C · provenance grade P1 · direction Not supported · retail band 5, Evidence says no. Derived by rule 1 - direction Not supported. Group: Pharmaceuticals, nootropics and controlled substances. Scope qualifier: HEADLINE OUTCOME IS SELF-REPORTED WELLBEING; CAVANNA'S OBJECTIVE COGNITIVE BATTERY WAS ALSO NULL OR ADVERSE against placebo; apparent benefits are attributable to expectancy

Best available evidence: Szigeti et al., eLife 2021 — self-blinding citizen-science placebo-controlled study; and Cavanna et al., Transl Psychiatry 2022;12:307 — randomised double-blind placebo-controlled within-subjects laboratory study, NCT05160220

Sample and design: Szigeti: 1,630 signed up, 240 started, 191 completed, 159 analysed at week 9; randomised to placebo, half-half or microdose arms under a self-blinding protocol; substances used were LSD (n = 147, 61%), an LSD analogue (n = 33, 14%) and psilocybin mushrooms (n = 57, 24%) (eLife 62878). Cavanna: 34 participants (11 female, age 31.26 ± 4.41 years), randomised double-blind placebo-controlled within-subjects; 0.5 g dried Psilocybe cubensis administered twice per active week (1 g total), assayed at 640.2 µg/g psilocybin, "estimated effective psilocybin dose approximately 0.9 mg" (PMC9346139)

Effect as reported: The effect does not survive placebo control. Szigeti week-5 placebo-versus-microdose adjusted treatment differences ± 95% CI: replicable psychological wellbeing 2.5 ± 5.6, p = 0.37; mindfulness (CAMS) 0.8 ± 1.5, p = 0.32; paranoia (GPTS) −1.6 ± 2.5, p = 0.21; satisfaction with life 0.4 ± 1.7, p = 0.67; Big-5 intellect −0.2 ± 1.2, p = 0.80; openness 0.3 ± 1.2, p = 0.57; neuroticism −0.3 ± 1.4, p = 0.70; extraversion −0.2 ± 1.2, p = 0.81; agreeableness 0.5 ± 1.1, p = 0.37; conscientiousness 0.8 ± 1.3, p = 0.24 — the authors conclude "our results also suggest that these improvements are not due to the pharmacological action of microdosing, but are rather explained by the placebo effect (lack of significant between-groups differences)" (eLife 62878). Cavanna: "For all other measurements there was no effect of microdosing except for few small changes towards cognitive impairment"; "0.5 g of dried mushroom material did not present significantly positive impact on creativity (divergent and convergent thinking), cognition, physical activity levels, and self-reported measures of mental health and well-being"; "we observed a trend towards impaired performance in some cognitive tasks (i.e., attentional blink and Stroop)"; "We conclude that expectation underlies at least some of the anecdotal benefits attributed to microdosing with psilocybin mushrooms" (PMC9346139)

Regulatory position: "Psilocybin is a Schedule I substance under the Controlled Substances Act", "meaning that it has a high potential for abuse, no currently accepted medical use in treatment in the United States, and a lack of accepted safety for use under medical supervision" (DEA Drug Fact Sheet). No FDA- or EMA-approved psilocybin medicine was identified. Non-US national scheduling: not checked

Safety signal: Cavanna observed "a trend towards impaired performance in some cognitive tasks (i.e., attentional blink and Stroop)" and "few small changes towards cognitive impairment" at the doses administered (PMC9346139). Schedule I status carries the DEA's finding of "a lack of accepted safety for use under medical supervision" (DEA). Quantified serious adverse events in microdosing trials: n.a.

Sources: Szigeti et al., eLife 2021 Cavanna et al., PMC9346139 ClinicalTrials.gov NCT05160220 DEA psilocybin fact sheet

Caffeine (1,3,7-trimethylxanthine)

It lets trained and regularly active adults keep going longer, and go faster over a set distance.

Claim as graded “Improves endurance exercise performance (time to exhaustion and time-trial speed) in trained and recreationally trained adults”

Evidence tier A · provenance grade P3 · direction Supported · retail band 1, Strong evidence. Derived by rule 6 - tier A/B with provenance P3+. Group: Ergogenic supplements.

Best available evidence: Umbrella review of 21 meta-analyses (BJSM 2020); supporting SR/MA of endurance running (PMC9824573) and multivariate meta-analysis of closed-end 45 s–8 min tests (Christensen 2017, PMC5422435)

Sample and design: 21 RCTs, 254 participants (220 men, 19 women, 15 unspecified) in the running meta-analysis (PMC9824573); 9 studies, 97 subjects, 12 performance tests in Christensen (PMC5422435); umbrella review covered 11 reviews / 21 meta-analyses, median 19 primary studies each (BJSM 2020)

Effect as reported: Time to exhaustion, running: Hedges g = 0.392, 95% CI 0.214 to 0.571, p < 0.001; time trials: g = −0.101, 95% CI −0.190 to −0.012, p = 0.026 (PMC9824573). Closed-end 45 s–8 min speed: Cohen's d = 0.41, 95% CI 0.15 to 0.68, p = 0.002 (PMC5422435). Aerobic-endurance effect sizes across the umbrella review ranged 0.22–0.61 (BJSM 2020)

Regulatory position: No authorised EU health claim. Caffeine performance claims are listed Non-authorised: "Improves physical performance" / "Enhances physical performance… can delay the onset of fatigue" (entries 737, 1486, 1489, EFSA opinion 2011;9(4):2053), and the caffeine + carbohydrate endurance claim (entry 543, EFSA 2011;9(6):2247) (EU Register). EFSA assessed exactly these claims — short-term high-intensity performance (737, 1486, 1489), endurance performance (737, 1486), endurance capacity (1488) and reduced perceived exertion (1488, 1490) — in opinion 2011;9(4):2053 (EFSA Journal 2053); the Commission did not authorise them. WADA: caffeine is in the 2026 Monitoring Program and is "not considered" a Prohibited Substance (WADA)

Safety signal: Fatal caffeine intoxication is documented: a PRISMA systematic review of 36 studies identified 92 deaths in which caffeine was the only cause of death, most attributed to ventricular fibrillation; among 5 caffeine-related deaths in athletes all were cardiac arrest from ventricular fibrillation (Cappelletti 2018, PMC5986491). EFSA safe intake: up to 400 mg/day for healthy adults, single doses up to 200 mg, 200 mg/day in pregnancy (EFSA)

Sources: BJSM umbrella PMC9824573 PMC5422435 EU Register EFSA 2053 EFSA caffeine safety WADA PMC5986491

Beta-alanine (β-alanine; carnosine precursor)

It lets healthy adults do more during hard efforts lasting roughly one to four minutes.

Claim as graded “Increases high-intensity exercise capacity in healthy adults during efforts of roughly 1–4 minutes”

Evidence tier B · provenance grade P3 · direction Supported · retail band 1, Strong evidence. Derived by rule 6 - tier A/B with provenance P3+. Group: Ergogenic supplements.

Best available evidence: Systematic review and 3-level mixed-effects meta-analysis (Saunders, BJSM 2017); contrasted with the closed-end 45 s–8 min meta-analysis (Christensen 2017, PMC5422435)

Sample and design: 40 studies, 65 exercise protocols, 70 exercise measures, 1461 participants (BA n = 746; placebo n = 715), double-blind placebo-controlled only (BJSM 2017); Christensen pooled 7 studies, 72 subjects, 11 tests (PMC5422435)

Effect as reported: Overall ES = 0.180, 95% CI 0.078 to 0.284; sensitivity analysis 0.180, 95% CI 0.076 to 0.292; greatest gains in the 0.5–10 min window, capacity outcomes larger than performance outcomes, smaller effect sizes in trained than non-trained participants (BJSM 2017). In closed-end 45 s–8 min performance tests the pooled effect was null: d = 0.17, 95% CI −0.12 to 0.46, p = 0.24 (PMC5422435)

Regulatory position: No authorised EU claim. Every beta-alanine entry in the EU Register is Non-authorised, including "Beta-alanine supplementation improves exercise performance" (1453), "…improves muscle work capacity" (1456), "…increases muscle buffering capacity" (1459), and "Delay onset of fatigue / improves short-duration high intensity performance" (436), all against EFSA opinion 2010;8(10):1729 (EU Register). WADA status not stated on the fetched Prohibited List page (WADA)

Safety signal: Paraesthesia (tingling) is described as the only reported side effect in the ISSN position stand (PMC4501114); no serious harm signal found

Sources: BJSM 2017 PMC5422435 PMC4501114 EU Register

Sodium bicarbonate (NaHCO₃)

It lets healthy adults perform better during hard efforts lasting roughly one to ten minutes.

Claim as graded “Improves high-intensity exercise performance in healthy adults during efforts of roughly 1–10 minutes”

Evidence tier A · provenance grade P3 · direction Supported · retail band 1, Strong evidence. Derived by rule 6 - tier A/B with provenance P3+. Group: Ergogenic supplements.

Best available evidence: Multivariate meta-analysis of closed-end performance tests lasting 45 s to 8 min (Christensen 2017, PMC5422435); umbrella review of 8 meta-analyses (Grgic 2021, JISSN)

Sample and design: 25 studies, 235 subjects, 33 performance tests (PMC5422435). Umbrella review: 8 meta-analyses, 5–26 studies each (average 13), 46–241 participants each; overall participant total not reported; 77–100% male (JISSN 2021)

Effect as reported: d = 0.40, 95% CI 0.27 to 0.54, p < 0.001 for average speed in 45 s–8 min closed-end tests (PMC5422435). Umbrella review: pooled ES 0.40 for endurance events ~45 s–8 min (25 studies) and 0.37 for muscle endurance (12 studies), with 95% CIs not reported for most outcomes; muscle strength ES −0.03 (JISSN 2021)

Regulatory position: No authorised EU claim. Sodium bicarbonate / bicarbonate entries are Non-authorised, including "can delay tiring of muscles in endurance sports" (entry 1403, EFSA 2011;9(6):2247) and "Sodium bicarbonate can reduce acid levels in the stomach" (entry 1653, EFSA 2010;8(2):1472) (EU Register). WADA status not stated on the fetched Prohibited List page (WADA)

Safety signal: Gastrointestinal side effects are documented and can abolish benefit: belching, bloating, nausea, vomiting, abdominal pain and osmotic diarrhoea; in one cited trial 4 of 21 participants were excluded for gastrointestinal discomfort, and in another 2 participants reported stomachache and diarrhoea (ISSN position stand)

Sources: PMC5422435 JISSN umbrella ISSN position stand EU Register

Dietary nitrate (beetroot juice; NO₃⁻)

It lets trained adults keep going longer.

Claim as graded “Improves endurance performance in trained adults”

Evidence tier A · provenance grade P3 · direction Contested · retail band 4, Experts disagree. Derived by rule 2 - direction Contested. Group: Ergogenic supplements.

Best available evidence: Umbrella review of systematic reviews with meta-analyses plus re-pooled primary RCTs (PMC12106159); supporting SR/MA of 73 RCTs (JISSN 2021) and the closed-end 45 s–8 min meta-analysis (PMC5422435)

Sample and design: 20 systematic reviews with meta-analyses, 180 primary studies, 2672 unique participants (PMC12106159); 73 RCTs, n = 1061 (JISSN 2021); 5 studies, 66 subjects, 8 tests (PMC5422435)

Effect as reported: Time to exhaustion: SMD 0.33, 95% CI 0.19 to 0.47, p < 0.001 (k = 41); time trial: SMD −0.03, 95% CI −0.14 to 0.09, p = 0.65 (k = 42); VO₂max: SMD −0.10, 95% CI −0.26 to 0.05; muscular endurance SMD 0.48, 95% CI 0.23 to 0.74 (PMC12106159). 73-RCT review: time to exhaustion MD 25.27 s, 95% CI 12.69 to 37.84 (low-quality evidence); time-trial performance MD −1.98 s, 95% CI −4.37 to 0.41, p = 0.1 (JISSN 2021). Closed-end 45 s–8 min: d = 0.19, 95% CI −0.03 to 0.40, p = 0.09, with the authors stating nitrate "appears most relevant for non-elite athletes or athletes with modest aerobic power" (PMC5422435)

Regulatory position: No authorised EU claim: no nitrate entry exists in the EU Register, and all beetroot (Beta vulgaris) entries relate to digestion, gut flora and immune function and are Non-authorised (entries 2401, 3072, 3073, 3074) (EU Register). WADA status not stated on the fetched Prohibited List page (WADA)

Safety signal: No serious harm signal found for performance doses. EFSA's acceptable daily intake for nitrate is 3.7 mg/kg body weight/day, and EFSA concluded nitrates added to food at permitted levels are safe (EFSA). Adverse gastrointestinal effects with some formulations were flagged by the 73-RCT review (JISSN 2021)

Sources: PMC12106159 JISSN 2021 PMC5422435 EU Register EFSA nitrate

Citrulline malate (CitMal)

Taken shortly before lifting, it lets strength-trained adults get more reps out of a session.

Claim as graded “Increases resistance-exercise volume (repetitions to failure) in strength-trained adults after acute pre-exercise ingestion”

Evidence tier B · provenance grade P3 · direction Contested · retail band 4, Experts disagree. Derived by rule 2 - direction Contested. Group: Ergogenic supplements.

Best available evidence: Systematic review and random-effects meta-analysis of double-blind placebo-controlled crossover RCTs (Vårvik 2021, IJSNEM; PDF)

Sample and design: 8 studies, 137 participants (111 men, 26 women; all but 9 had ≥6 months strength-training experience), 14 exercises; studies used 6–8 g CitMal 40–60 min pre-exercise (IJSNEM 2021)

Effect as reported: SMD (Hedges' g) = 0.196, 95% CI 0.029 to 0.364, p = .022; weighted mean difference 3 ± 5 repetitions (6.4 ± 7.9%). Subgroups were null: lower body SMD 0.266, 95% CI −0.001 to 0.533, p = .051; upper body SMD 0.166, 95% CI −0.050 to 0.382, p = .131. Trim-and-fill correction for funnel asymmetry reduced the estimate to a non-significant SMD 0.104, 95% CI −0.017 to 0.216; removing one study moved the main result to p = .052 (IJSNEM PDF)

Regulatory position: No authorised EU claim; no citrulline or L-citrulline entry exists in the EU Register at all (EU Register). WADA status not stated on the fetched Prohibited List page (WADA)

Safety signal: None found

Sources: IJSNEM 2021 IJSNEM PDF PMC8571142 critical review EU Register

HMB (β-hydroxy-β-methylbutyrate)

It builds more muscle in weight-training adults aged 18 to 50 than the training alone would.

Claim as graded “Increases lean body mass (fat-free mass) in resistance-training adults aged 18–50 beyond training alone”

Evidence tier B · provenance grade P3 · direction Not supported · retail band 5, Evidence says no. Derived by rule 1 - direction Not supported. Group: Ergogenic supplements.

Best available evidence: Cochrane-methodology systematic review and random-effects meta-analysis of double-blind RCTs (Jakubowski 2020, PMC7285233)

Sample and design: 11 double-blind RCTs in the meta-analysis; 302 participants in the body-composition analysis and 248 in the strength analysis; mean age 27 years; interventions lasted 7.6 ± 4.0 weeks; studies used 3.0 g HMB/day (PMC7285233)

Effect as reported: Fat-free mass: MD 0.29 kg, 95% CI −0.01 to 0.60 kg, p = 0.06 — not significant. Total body mass MD 0.34 kg, 95% CI 0.03 to 0.66 kg, p < 0.05; fat mass MD 0.10 kg, 95% CI −0.42 to 0.23 kg, p = 0.57 (PMC7285233)

Regulatory position: No authorised EU claim. All HMB entries are Non-authorised, including "HMB can increase gains in lean body mass during resistance training" (entry 1582) and "HMB has been shown to increase lean muscle mass" (entry 1579), both against EFSA opinion 2011;9(6):2227 (EU Register). WADA status not stated on the fetched Prohibited List page (WADA)

Safety signal: None found

Sources: PMC7285233 EU Register

Whey protein (protein supplementation)

It adds muscle in adults who lift weights.

Claim as graded “Increases lean body mass / fat-free mass gain in adults performing resistance training”

Evidence tier A · provenance grade P3 · direction Supported · retail band 1, Strong evidence. Derived by rule 6 - tier A/B with provenance P3+. Group: Ergogenic supplements.

Best available evidence: Systematic review, meta-analysis and meta-regression of RCTs (Morton, BJSM 2018); three-level meta-analysis of protein-intake RCTs (Tagawa 2022, PMC8978023); whey-specific isocaloric meta-analysis (Castro 2019, PMC6769754)

Sample and design: 49 RCTs, 1863 participants, mean age 35 ± 20 years, RET ≥6 weeks (BJSM 2018); lean-body-mass analysis of 66 studies / 93 intervention groups / 2665 subjects (PMC8978023); whey-only isocaloric analysis of 8 RCTs, 246 participants, all men (PMC6769754)

Effect as reported: Fat-free mass: MD 0.30 kg, 95% CI 0.09 to 0.52 kg, p = 0.007; 1RM strength MD 2.49 kg, 95% CI 0.64 to 4.33 kg, p = 0.01; total body mass MD 0.11 kg, 95% CI −0.23 to 0.46, p = 0.52 (BJSM 2018). Lean body mass SMD 0.22, 95% CI 0.15 to 0.29, p < 0.01, moderate certainty, ≈0.5–0.7 kg between-group difference (PMC8978023). Whey vs isocaloric carbohydrate placebo alone: FFM WMD 0.26, 95% CI −0.32 to 0.83, p = 0.381 — non-significant (PMC6769754)

Regulatory position: Authorised generic claims exist for protein: "Protein contributes to a growth in muscle mass" and "Protein contributes to the maintenance of muscle mass" (entries 415, 417, 593, 594, 595, 715, 1398; EFSA 2010;8(10):1811 and 2011;9(6):2203), authorised by Commission Regulation (EU) 432/2012 of 16/05/2012. The whey-specific claim "Consumption of whey in conjunction with resistance exercise supports an increase in lean body mass and strength" (entry 429, EFSA 2010;8(10):1818) is Non-authorised (EU Register)

Safety signal: None found

Sources: BJSM 2018 PMC8978023 PMC6769754 EU Register

BCAAs (branched-chain amino acids: leucine, isoleucine, valine)

Taken on their own, they build muscle in adults who already eat enough protein.

Claim as graded “Isolated BCAA supplementation increases muscle protein synthesis or lean mass in adults already consuming adequate dietary protein”

Evidence tier D · provenance grade P0 · direction Not supported · retail band 5, Evidence says no. Derived by rule 1 - direction Not supported. Group: Ergogenic supplements.

Best available evidence: Biochemical review of the human literature (Wolfe 2017, PMC5568273); PRISMA systematic review of pure-BCAA RCTs in which pooling was judged impossible (Julea & Saleh 2025, PMC12674588)

Sample and design: For the muscle-protein-synthesis claim: "no studies in human subjects in which the response of muscle protein synthesis to orally-ingested BCAAs alone was quantified", and only 2 intravenous-infusion studies (10 and 8 subjects) (PMC5568273). For body composition: 22 RCTs, 511 participants, 5–100 per study; no meta-analysis was performed because all included RCTs had unclear/high risk of bias and there was substantial heterogeneity (PMC12674588)

Effect as reported: no numeric effect reported for the pooled claim. The two intravenous BCAA studies found muscle protein synthesis and breakdown fell by the same degree with the balance remaining negative; a theoretical ceiling of about a 15% rise in synthesis (fractional synthetic rate ≈0.050%/h to ≈0.057%/h) was described as an over-estimate (PMC5568273). Individual body-composition RCTs were inconsistent: no between-group difference in Areces 2014 (n = 46, p = 0.13) or Bagheri 2021 (n = 30), versus muscle mass +2.2 ± 1.3 kg vs −0.6 ± 2.3 kg, p < 0.001 in Muscella 2024 (n = 100) (PMC12674588)

Regulatory position: No authorised EU claim. Every BCAA entry in the EU Register is Non-authorised, including "BCAAs increase protein synthesis, a vital part of the muscle-building process" (entry 444), "BCAAs support muscle growth" (442, 444) and "Improvement of muscle protein synthesis" (451), all against EFSA opinion 2010;8(10):1790 (EU Register). WADA status not stated on the fetched Prohibited List page (WADA)

Safety signal: None found

Sources: PMC5568273 PMC12674588 EU Register

L-citrulline / L-arginine

They open up blood flow and so let healthy adults perform better.

Claim as graded “Increase nitric-oxide-mediated blood flow and thereby improve exercise performance in healthy adults”

Evidence tier B · provenance grade P2 · direction Contested · retail band 4, Experts disagree. Derived by rule 2 - direction Contested. Group: Ergogenic supplements.

Best available evidence: Systematic review and random-effects meta-analysis of arginine and performance (Viribay 2020, PMC7282262); fixed-effects meta-analysis of citrulline and endurance (JISSN 2023); direct blood-flow RCT (PMC5999519)

Sample and design: Arginine: 18 studies in the systematic review, 15 in the meta-analysis, 394 participants (386 men, 8 women; 282 trained or elite) (PMC7282262). Citrulline: 9 RCTs, 158 participants, crossover, all but one double-blind (JISSN 2023). Blood flow measured directly in 1 randomised double-blind crossover trial of 25 older adults (13 women, 12 men) using 6 g/day L-citrulline for 14 days (PMC5999519)

Effect as reported: Arginine, aerobic performance: Hedges' g = 0.84, 95% CI 0.12 to 1.56, p = 0.02 with I² = 89%; anaerobic performance g = 0.24, 95% CI 0.05 to 0.43, p = 0.01, I² = 0% (PMC7282262). Citrulline, endurance: time to exhaustion SMD 0.03, 95% CI −0.27 to 0.33, p = 0.83; time to completion SMD −0.07, 95% CI −0.39 to 0.26, p = 0.67 — both null (JISSN 2023). Blood flow: femoral blood flow 521 ± 134 vs 584 ± 166 mL/min in men only, p = 0.04, but "the difference following L-citrulline was removed" after adjusting for baseline diastolic blood pressure (p = 0.10), and no change in women (PMC5999519)

Regulatory position: No authorised EU claim. No citrulline entry exists in the EU Register; every arginine entry is Non-authorised, including "Support of normal blood circulation" against the health relationship "Improvement of endothelium-dependent vasodilation" (entries 1443, 664) and "Power for muscles / Increases nitric oxide production" (entry 1820), all against EFSA opinion 2011;9(4):2051 (EU Register). WADA status not stated on the fetched Prohibited List page (WADA)

Safety signal: None found

Sources: PMC7282262 JISSN 2023 PMC5999519 PMC10005484 narrative review EU Register

Glutamine (L-glutamine)

It speeds recovery, strengthens immunity or changes body composition in adult athletes.

Claim as graded “Improves recovery, immune function or body composition in adult athletes”

Evidence tier B · provenance grade P3 · direction Not supported · retail band 5, Evidence says no. Derived by rule 1 - direction Not supported. Group: Ergogenic supplements.

Best available evidence: Systematic review and meta-analysis of clinical trials, PROSPERO CRD16038438 (Ramezani Ahmadi 2019, Clin Nutr; full record)

Sample and design: 47 studies in the systematic review, 25 trials in the meta-analysis (20 on immune function, 6 on body composition); total participants not stated in the fetched pages (PubMed; ScienceDirect)

Effect as reported: Weight: WMD −1.36, 95% CI −2.55 to −0.16, p = 0.02. Neutrophil count at doses above 200 mg/kg: WMD −605.77, 95% CI −1200.0 to 52.1, P = 0.03. Post-exercise blood glucose with glutamine dipeptide: WMD 0.51 mmol/L, 95% CI 0.18 to 0.83, P = 0.002. No association with any other outcome; the authors concluded "glutamine supplementation has no effect on athletics immune system, aerobic performance, and body composition" (PubMed)

Regulatory position: No authorised EU claim. All glutamine entries are Non-authorised, including "Glutamine is considered essential for repair and recovery / Glutamine can aid in muscle tissue repair" against "skeletal muscle tissue repair" (entry 721) and the muscle-mass claim (entry 719), against EFSA opinion 2011;9(6):2225 (EU Register). WADA status not stated on the fetched Prohibited List page (WADA)

Safety signal: None found

Sources: PubMed 29784526 ScienceDirect record EU Register

Betaine anhydrous (trimethylglycine)

Taken for at least a week, it makes trained adults stronger at their maximum.

Claim as graded “Increases maximal muscular strength (1RM/isokinetic/isometric force) in healthy trained adults taking it for at least 7 days”

Evidence tier B · provenance grade P3 · direction Contested · retail band 4, Experts disagree. Derived by rule 2 - direction Contested. Group: Ergogenic supplements.

Best available evidence: Systematic review and random-effects meta-analysis of double-blind placebo-controlled trials (J Sports Sci 2024, PubMed 39514262; full text PDF); body-composition counterpart (Br J Nutr 2022, PubMed 34743773)

Sample and design: 17 studies, 317 participants (79% male n = 250, 21% female n = 67); 15 of 17 randomised, all double-blind and placebo-controlled; 9 studies / 17 outcomes entered the maximal-strength analysis; supplementation lasted 7 to 98 days (mean 26.2 ± 23.4 days) and studies administered 2.0–6.0 g/day, most commonly 2.5 g (J Sports Sci PDF)

Effect as reported: Maximal strength SMD = 0.47, 95% CI 0.04 to 0.89, p = 0.03, I² = 78%; lower body SMD = 0.49, 95% CI 0.01 to 0.98, p = 0.05, I² = 69%; upper body non-significant, SMD = 0.50, 95% CI −0.25 to 1.25, p = 0.19; vertical jump SMD = 0.36, 95% CI 0.03 to 0.69 after excluding one low-quality study; "no significant effects were found for upper body strength, cycling sprint power, bench press throws power, or muscular endurance". Leave-one-out analyses removed the lower-body effect three separate times (SMD 0.50, 95% CI −0.05 to 1.05, p = 0.07; SMD 0.49, 95% CI −0.06 to 1.04, p = 0.08 twice) (J Sports Sci PDF). Body composition is null: body mass WMD −0.40 kg, 95% CI −1.46 to 0.64, p = 0.447; fat mass WMD −0.57 kg, 95% CI −2.14 to 0.99, p = 0.473; fat-free mass WMD 0.61 kg, 95% CI −1.27 to 2.49, p = 0.527 (Br J Nutr 2022)

Regulatory position: The only authorised betaine claim in the EU is metabolic, not ergogenic: "Betaine contributes to normal homocysteine metabolism" (entry 4325, EFSA 2011;9(4):2052, Commission Regulation (EU) 432/2012 of 16/05/2012). No authorised or pending strength, muscle or performance claim for betaine appears in the Register (EU Register)

Safety signal: The authorised EU claim carries a mandatory consumer warning in its conditions of use: "a daily intake in excess of 4 g may significantly increase blood cholesterol levels" (EU Register, entry 4325)

Sources: PubMed 39514262 J Sports Sci full text PDF PubMed 34743773 EU Register

Taurine

A single dose lets healthy adults keep going longer.

Claim as graded “Improves endurance exercise performance (time to exhaustion / overall exercise performance) in healthy adults after acute ingestion”

Evidence tier B · provenance grade P3 · direction Supported · retail band 1, Strong evidence. Derived by rule 6 - tier A/B with provenance P3+. Group: Ergogenic supplements. Scope qualifier: low to very low GRADE certainty

Best available evidence: Three-level random-effects systematic review and meta-analysis of randomised trials, GRADE-appraised (Scand J Med Sci Sports 2025, PubMed 40852891); earlier meta-analysis with meta-regression (Sports Med 2018, PubMed 29546641)

Sample and design: 23 eligible randomised trials, k = 69 effects, N = 308 participants, six databases searched to July 2025 (PubMed 40852891); 10 peer-reviewed articles with a 7-trial time-to-exhaustion sub-analysis, doses of 1–6 g/day administered as single doses and for up to 2 weeks (PubMed 29546641)

Effect as reported: Overall exercise performance g = 0.25, 95% CI 0.10 to 0.39, I² = 61%; p-value not stated on the record; "the overall certainty of evidence, assessed using GRADE, was rated low to very low due to heterogeneity, imprecision, and risk of bias", several subgroup effects were "attenuated after excluding influential studies" and "prediction intervals frequently included the null" (PubMed 40852891). Earlier pooling: endurance performance Hedges' g = 0.40, 95% CI 0.12 to 0.67, P = 0.004; time-to-exhaustion trials g = 0.43, 95% CI 0.12 to 0.75, P = 0.007; dose did not moderate the effect (P > 0.05) (PubMed 29546641)

Regulatory position: No authorised EU claim. Every taurine entry in the Register is Non-authorised, including the exact marketed wording "Helps to delay the onset of fatigue… enhances endurance and helps to maintain peak effort during times of high physical demand" against the health relationship "delay in the onset of fatigue and enhancement of physical performance" (entry 1660, EFSA 2009;7(9):1260), and "Helps to enhance tonus and vitality" against "delay in the onset of physical fatigue during exercise" (entry 1958, EFSA 2011;9(4):2035) (EU Register)

Safety signal: None found. EFSA's ANS Panel concluded that "exposure to taurine and d-glucuronolactone through regular consumption of energy drinks was not of safety concern" and confirmed a NOAEL of 1,000 mg per kilogram of bodyweight per day for taurine (EFSA)

Sources: PubMed 40852891 PubMed 29546641 EFSA news EU Register

Alpha-GPC (L-alpha-glycerylphosphorylcholine; choline alfoscerate)

A single dose sharpens focus and thinking in healthy adults.

Claim as graded “Sharpens focus and cognitive performance in healthy adults after an acute dose”

Evidence tier C · provenance grade P3 · direction Insufficient · retail band 6, Too little evidence yet. Derived by rule 3 - direction Insufficient. Group: Ergogenic supplements. Scope qualifier: healthy adults

Best available evidence: No pooled evidence exists in healthy adults. Strongest single source in the marketed population is one randomised, double-blind, placebo-controlled three-way crossover RCT (Nutrients 2024, PMC11644786). The only meta-analysis is in a different population — adult-onset cognitive impairment (PROSPERO CRD42022356965, PMC10041421)

Sample and design: Healthy adults: 21 recruited, n = 20 analysed, healthy resistance-trained men aged 31.3 ± 11.0 years, Latin-square crossover, retrospectively registered as NCT06690619; single doses of 315 mg and 630 mg A-GPC and a resistant-dextrin placebo were administered (PMC11644786). Clinical population: 8 studies (7 RCTs + 1 prospective cohort), 861 participants (433 intervention / 428 control), published 1993–2022, 1,200 mg/day administered for 90–720 days (PMC10041421)

Effect as reported: Healthy men, Stroop Total Score change: 630 mg 13.0 ± 8.2 vs placebo 5.2 ± 9.0, p = 0.013, Cohen's d = 0.61; 315 mg 10.8 ± 7.7 vs placebo, p = 0.046, d = 0.48; Stroop time per score −0.12 ± 0.09 s vs −0.05 ± 0.09 s, p = 0.021, d = 0.56; "No significant differences between groups were found for the Flanker and N-Back" assessments (PMC11644786). In Alzheimer's disease, α-GPC vs placebo or comparator: MMSE MD = 3.50, 95% CI 0.36 to 6.63 (3 RCTs, n = 449); as add-on to donepezil MMSE MD = 1.72, 95% CI 0.20 to 3.25 and ADAS-Cog MD = −5.76, 95% CI −8.07 to −3.46 (n = 350); p-values not reported on the fetched page (PMC10041421)

Regulatory position: No EU health claim exists for alpha-GPC or choline alfoscerate — the substance does not appear in the EU Register at all. The related generic claims for choline that were authorised concern lipid metabolism, homocysteine metabolism and liver function (entries 3090, 3186, 712/1633, EFSA 2011;9(4):2056), while the cognitive wording "Choline supports cognitive functioning. Choline helps maintain memory and brain function" is Non-authorised (entry 1502) (EU Register). Choline alfoscerate is simultaneously a nationally authorised medicinal product in EU Member States — EMA's PSUSA/00010599/202208 list names DELECIT and GLIATILIN products authorised in Italy (IT) and Poland (PL) (EMA)

Safety signal: Population-based retrospective cohort of 12,008,977 South Korean adults aged ≥50 (108,877 α-GPC users) found α-GPC use associated with higher 10-year stroke risk: total stroke aHR 1.46, 95% CI 1.43–1.48; ischaemic 1.36, 95% CI 1.33–1.39; haemorrhagic 1.36, 95% CI 1.28–1.44; matched cohort total stroke aHR 1.43, 95% CI 1.41–1.46; risk rose "in a dose-response manner" (JAMA Netw Open 2021, PMC8613599; PubMed 34817582)

Sources: PMC11644786 PMC10041421 PMC8613599 PubMed 34817582 EMA PSUSA list EU Register

Ecdysterone (beta-ecdysterone / 20-hydroxyecdysone, from Cyanotis or spinach extract)

It builds muscle and adds bench-press strength in young men who lift weights.

Claim as graded “Builds muscle mass and increases bench-press strength in resistance-training young men”

Evidence tier C · provenance grade P3 · direction Contested · retail band 4, Experts disagree. Derived by rule 2 - direction Contested. Group: Ergogenic supplements.

Best available evidence: No meta-analysis exists. Best single source is one 10-week controlled intervention study (Isenmann 2019, Arch Toxicol, PubMed 31123801; publisher record); contradicted by an 8-week double-blind placebo-controlled RCT (Wilborn 2006, PMC2129166); a separate 12-week double-blind placebo-controlled RCT of a spinach extract in older adults (Pérez-Piñero 2021, PMC8706266)

Sample and design: Isenmann: 46 young men (mean age 25.6 ± 3.7 y), 10 weeks of strength training; 20 men took 200 mg/day 20-hydroxyecdysone, 10 took 800 mg/day, 12 received placebo, plus a 12-volunteer non-training control on 200 mg — capsules contained 6 mg ecdysterone plus 100 mg leucine each (PubMed 31123801; group breakdown and capsule composition per PMC11085066 and PMC8706266). Wilborn: 45 resistance-trained males, double-blind placebo-controlled parallel groups matched on fat-free mass, 8 weeks; the ecdysterone arm received 100 mg/day of Polypodium vulgare/suma root standardised for 30 mg of 20-hydroxyecdysone (PMC2129166). Pérez-Piñero: 51 randomised, 45 completing (23 spinach extract / 22 placebo), men and postmenopausal women aged 50–75, 2 g/day of spinach extract for 12 weeks with supervised training (PMC8706266)

Effect as reported: no numeric effect reported on any page fetched for the headline claim. Isenmann states only that "significantly higher increases in muscle mass were observed in those participants that were dosed with ecdysterone" and "significantly more pronounced increases in one-repetition bench press performance were observed" — no means, no confidence intervals and no p-values appear on the abstract or publisher record, and the full text could not be retrieved (PubMed 31123801; Springer). Wilborn, by contrast: "No changes were observed in training adaptation and in anabolic/catabolic effect" (PMC2129166; PMC11085066). Spinach extract in over-50s beat placebo on knee-extension peak torque at 180° s⁻¹ (72.4 ± 19.3 vs 64.2 ± 12.9 Nm, time × product interaction p = 0.002) and isometric peak torque (p = 0.005), but not on maximal dynamic force 1RM (60.1 ± 16.7 vs 57.3 ± 15.1 kg, p = 0.729) or handgrip strength (right hand p = 0.449) (PMC8706266)

Regulatory position: Ecdysterone as an isolated substance has no EU health claim and no entry in the EU Register (EU Register). A peer-reviewed regulatory review states that "ecdysterone (or other ecdysteroids) as isolated substance was not mentioned in the RASFF panel, likewise in EU Novel Food Catalogue" and that "ecdysterone not include in the EFSA Novel food catalogue", while Rhaponticum carthamoides, one plant source, is catalogued as non-novel; the same review notes that "EFSA Emerging Risks Exchange Network (EREN) briefed a note on possible emerging health risks, concerns the use of ecdysterone in food supplements for anabolic purposes" (Food Reviews International 2023)

Safety signal: None found in the fetched trials: Isenmann reported "no increase in biomarkers for liver or kidney toxicity was noticed" (PubMed 31123801), and a review of the same trial records no adverse impact on creatinine, GGT, GOT or GPT and an unaffected steroid profile (PMC11085066)

Sources: PubMed 31123801 Springer record PMC2129166 PMC8706266 PMC11085066 Food Reviews International 2023 EU Register

p-Synephrine (Citrus aurantium / bitter orange)

It burns fat and causes weight loss in overweight and healthy adults.

Claim as graded “Burns fat and produces weight loss in overweight and healthy adults”

Evidence tier B · provenance grade P3 · direction Not supported · retail band 5, Evidence says no. Derived by rule 1 - direction Not supported. Group: Ergogenic supplements.

Best available evidence: PRISMA systematic review and meta-analysis of placebo-controlled human clinical trials, PROSPERO 359626 (Nutrients 2022, PMC9572433; PubMed 36235672)

Sample and design: 18 double-blind trials (7 crossover, 5 parallel, 4 within-subjects/other, 2 counterbalanced), participants aged 18–51; 14 trials assessed acute effects and 4 assessed 4–8 weeks of treatment. Weight loss was pooled from 3 trials in 94 subjects; body fat percentage n = 94; fat mass n = 69; fat-free mass n = 78; blood pressure from 11 trials in 222 subjects and heart rate from 9 trials in 129 subjects; doses of 10–214 mg were administered (PMC9572433)

Effect as reported: Weight loss MD 0.60 kg, 95% CI −5.62 to 6.83, p = 0.85 (i.e. numerically favouring placebo); body fat −1.87%, 95% CI −3.92 to 0.18, p = 0.07; fat mass MD −0.32 kg, 95% CI −3.76 to 3.11, p = 0.85; fat-free mass MD 0.47 kg, 95% CI −4.19 to 5.13, p = 0.84; respiratory exchange ratio unchanged at 1 h (0.00, 95% CI −0.03 to 0.03, p = 0.91), 2 h (−0.02, 95% CI −0.12 to 0.09, p = 0.75) and 3 h (−0.02, 95% CI −0.12 to 0.08, p = 0.73) (PMC9572433)

Regulatory position: No authorised EU claim: searched entry-by-entry, neither "synephrine" nor "Citrus aurantium" appears anywhere in the EU Register; the only bitter-orange-adjacent entries are for other Citrus species and citrus bioflavonoids, all Non-authorised (e.g. entries 1471, 1799, 2025, 3667) (EU Register). FDA/EMA status: Not checked

Safety signal: Blood pressure rose significantly with prolonged use in the same meta-analysis: systolic +6.37 mmHg, 95% CI 1.02 to 11.72, p = 0.02 and diastolic +4.33 mmHg, 95% CI 0.48 to 8.18, p = 0.03 after 8 weeks at 10–49 mg/day (2 trials, 75 participants) (PMC9572433). A systematic review of case reports on synephrine-containing pre-workout supplements collected 30 case reports covering 35 patients aged 16–57 across 8 countries (mean synephrine dose 41.9 ± 28.8 mg), with syncope and collapse mentioned 6 times, dizziness 6 times, seizures 2 times and amnesia 2 times (Cardiovasc Toxicol 2023, PMC9859859)

Sources: PMC9572433 PubMed 36235672 PMC9859859 EU Register

Nicotine, non-smoked delivery (pouches, gum), used for performance or cognition

It sharpens attention and thinking in healthy adults who do not smoke.

Claim as graded “Sharpens attention and cognitive performance in healthy non-smoking adults”

Evidence tier B · provenance grade P3 · direction Supported · retail band 1, Strong evidence. Derived by rule 6 - tier A/B with provenance P3+. Group: Ergogenic supplements. Scope qualifier: patch and mixed-route delivery; no pooled pouch-specific data found

Best available evidence: Systematic review and meta-analysis of clinical trials of nicotine patches versus placebo in healthy non-smoking adults (Acta Neurol Scand 2021, PubMed 33899218); broader meta-analysis of acute nicotine and smoking effects covering gum, patch and other routes (Psychopharmacology 2010, PubMed 20414766)

Sample and design: 31 publications, 978 subjects, healthy non-smoking adults, nicotine patch vs placebo patch (PubMed 33899218); 41 double-blind placebo-controlled laboratory studies published 1994–2008 in "healthy adult nonsmokers or smokers who were not tobacco-deprived or minimally deprived (≤2 h)"; total participants not stated on the record (PubMed 20414766)

Effect as reported: Overall cognitive function SMD = 0.233, 95% CI 0.111 to 0.355, p < .001 (I² = 50.17%); attention SMD = 0.231, 95% CI 0.106 to 0.356, p < .001 (I² = 34.07%); memory SMD = 0.270, 95% CI −0.293 to 0.833, p = .347 — not significant, I² = 77.98% (PubMed 33899218). The 41-study meta-analysis found significant positive effects in six of nine domains — fine motor, alerting attention accuracy and response time, orienting attention RT, short-term episodic memory accuracy and working memory RT — with an "effect size range = 0.16 to 0.44"; per-domain confidence intervals and p-values are not stated on the record (PubMed 20414766)

Regulatory position: Not a food and therefore outside the EU Register, which contains no nicotine entry (EU Register). The European Commission stated on 25 October 2023 that "tobacco-free nicotine pouches are currently outside the scope of the Tobacco Products Directive 2014/40/EU (TPD)", while "current EU legislation prohibits the sale of tobacco products for oral use, like snus, except in Sweden"; Member State approaches range "from banning them completely — as in Belgium and the Netherlands — to allowing them to be sold without any restrictions whatsoever", and regulation of these products is under an ongoing TPD evaluation (European Parliament E-002498/23 and Commission answer002498_EN.pdf))

Safety signal: US poison-centre analysis of 134,663 unintentional single-substance nicotine ingestions in children under 6 (2010–2023) recorded 39 ingestions with major effects and 2 fatalities; 76.2% involved children under 2. The nicotine pouch ingestion rate rose 763.1% from 2020 to 2023 while other formulations declined (Pediatrics 2025, PubMed 40658193), and pouches "were also more likely to be associated with serious medical outcomes or hospital admissions than other nicotine products like gum/lozenges, e-liquids, powder/granules, and tablets/capsules/caplets" (Nationwide Children's)

Sources: PubMed 33899218 PubMed 20414766 PubMed 40658193 Nationwide Children's European Parliament / Commission answer EU Register

L-tyrosine

It holds mental performance steady in healthy adults under short-term stress such as cold, altitude or long shifts.

Claim as graded “Maintains mental performance and cognitive function in healthy adults under acute stress (cold, altitude, sustained operations)”

Evidence tier D · provenance grade P3 · direction Insufficient · retail band 6, Too little evidence yet. Derived by rule 3 - direction Insufficient. Group: Ergogenic supplements.

Best available evidence: Rapid evidence assessment of the literature using modified GRADE and SIGN 50 quality appraisal (Mil Med 2015, Oxford Academic); narrative review of the same literature (J Psychiatr Res 2015, PubMed 26424423)

Sample and design: 10 randomised controlled trials plus 4 controlled clinical trials met inclusion criteria out of 421 citations. GRADE domains: environmental stressors 175 participants across 6 studies; exercise 47 participants across 4 studies; the 4 CCTs covered 77 participants, 57 of them exposed to cold immersion or simulated altitude. "The majority (80%) of the 10 included RCTs were of poor (−) quality with only 2 high (+) quality studies" (Mil Med 2015)

Effect as reported: no numeric effect reported. The review states plainly: "Authors did not describe or report effect size for this review's outcomes of interest because of the lack of author reporting", magnitude of effect was graded "ND" (not described) for both the exercise and environmental-stressor domains, and confidence in the estimate was graded C in both. It issued a "Weak Recommendation in Favor" for environmental stressors and "No Recommendation" for exercise. Individual trials ran both ways: "TYR Does Not Significantly Affect Endurance, Muscle Strength or Anaerobic Power in Healthy Men" (n = 20) versus "TYR Improves Short-term Memory (p < 0.05)… Stroop Task (p < 0.03)" (n = 16) and improved reaction time (p < 0.01, n = 20) (Mil Med 2015). The narrative review concludes tyrosine "does seem to effectively enhance cognitive performance, particularly in short-term stressful and/or cognitively demanding situations… but only when neurotransmitter function is intact and DA and/or NE is temporarily depleted", and that its potential for enhancing physical exercise "seems minimal" (PubMed 26424423)

Regulatory position: No authorised EU claim. All tyrosine entries are Non-authorised: "Helps maintain physical and mental concentration in cases of temporary stress" (entry 1672) and "Tyrosine helps maintain mental focus and performance during exposure to environmentally adverse conditions. Tyrosine limits mental fatigue during exposure to environmentally adverse conditions" (entry 440), both against the health relationship "increased attention" and EFSA opinion 2011;9(6):2270; further Art.13(1) entries 1929 and 1930 are non-authorised, as is the Art.13(5) dossier "L-tyrosine and contribution to normal synthesis of dopamine" (Q-2011-00319, Commission Regulation (EU) No 375/2010 of 03/05/2010) (EU Register)

Sources: Mil Med 2015 PubMed 26424423 EU Register

Exogenous ketones (ketone monoesters / beta-hydroxybutyrate salts)

It lets trained athletes keep going longer.

Claim as graded “Improves endurance exercise performance in trained athletes”

Evidence tier B · provenance grade P3 · direction Not supported · retail band 5, Evidence says no. Derived by rule 1 - direction Not supported. Group: Ergogenic supplements.

Best available evidence: Two independent systematic reviews with meta-analysis of randomised trials (IJSNEM 2022, Human Kinetics; IJSPP 2020, PubMed 32045881)

Sample and design: IJSNEM: 8 studies, 80 individuals (77 men, 3 women), all crossover, all trained athletes aged 25–38, sample sizes 8–12, published 2016–2020; ketone doses of approximately 500 to 922 mg/kg body weight were administered (IJSNEM 2022). IJSPP: 13 RCTs meeting inclusion criteria; total participants not stated on the record (PubMed 32045881)

Effect as reported: Overall endurance performance Hedges' g = 0.136, 95% CI −0.195 to 0.467, p = .419 (I² = 56.888%); time to exhaustion g = −0.002, 95% CI −0.312 to 0.308, p = .989; time-trial completion g = 0.057, 95% CI −0.282 to 0.395, p = .744 (IJSNEM 2022). Independently: overall exercise performance g = −0.05, 95% CI −0.30 to 0.20, p = .68; endurance time trial g = −0.04, 95% CI −0.35 to 0.28, p = .82; ketone esters g = −0.07, 95% CI −0.38 to 0.24, p = .66; ketone salts g = −0.02, 95% CI −0.45 to 0.41, p = .93 — despite all studies confirming raised plasma ketone concentrations (PubMed 32045881)

Regulatory position: No EU health claim; no ketone, ketone ester or beta-hydroxybutyrate entry exists in the EU Register (EU Register). BHB salts are regulated as a novel food: EFSA's NDA Panel assessed sodium, magnesium and calcium BHB salts under Regulation (EU) 2015/2283 and concluded the applicant's identity, production and compositional data were "overall considered unsatisfactory", that "the Panel cannot establish a safe intake level of the NF", and that "the safety of the NF has not been established" (EFSA Journal 2022, PubMed 36254193)

Safety signal: EFSA concluded that the safety of BHB salts as a novel food "has not been established" and that no safe intake level could be set (PubMed 36254193). Gastrointestinal distress was higher on ketone than control in 6 of the 7 studies that assessed it — for example 9 of 11 participants symptomatic on ketone ester versus 4 of 11 on control (Evans & Egan 2018), and 5 of 8 versus 4 of 8 (Evans et al. 2019); the review cautions these data were not statistically analysed (IJSNEM 2022)

Sources: IJSNEM 2022 PubMed 32045881 PubMed 36254193 EU Register

Phosphatidic acid

It adds muscle in weight-trained men who also do endurance training.

Claim as graded “Increases lean body mass in resistance-trained men training concurrently”

Evidence tier D · provenance grade P3 · direction Not supported · retail band 5, Evidence says no. Derived by rule 1 - direction Not supported. Group: Ergogenic supplements.

Best available evidence: Scoping review of the whole human literature, explicitly downgraded from a systematic review because too few studies exist (J Sports Sci 2022, PubMed 34706625); the two underlying RCTs were fetched directly (Hoffman 2012, PMC3506449; Andre 2016, PMC4974867)

Sample and design: Scoping review: 2,009 articles retrieved, 6 studies analysed, published 2012–2019, five in adult male and one in elderly male populations; total participants not stated, and no pooling was performed — "Due to the small number of studies, this is a scoping review" (PubMed 34706625). Hoffman 2012: randomised, double-blind, placebo-controlled, 20 resistance-trained men enrolled and 16 completing (7 phosphatidic acid, 9 placebo), 750 mg/day administered for 8 weeks (PMC3506449). Andre 2016: double-blind randomised placebo-controlled, 32 resistance-trained men enrolled and 28 completing (375 mg n = 9, 250 mg n = 9, placebo n = 10), 8 weeks (PMC4974867)

Effect as reported: no pooled numeric effect exists. The scoping review reports that "three studies suggested no effect of PA on lean body mass, while the remaining showed a possible positive effect", that in one of those the supplement "included other potentially anabolic substances, precluding an isolated effect of PA", and concludes "the evidence does not support the supplementation with PA to increase performance or improve body composition in young or elderly men" (PubMed 34706625). Andre 2016 found a main effect of time but no group × time effect on lean mass (57.5 ± 5.9 kg on 375 mg, 63.0 ± 7.1 kg on 250 mg, 60.5 ± 9.1 kg on placebo at day 57; time p = .008, group × time p = .55) or lower-body strength (time p < .001, group × time p = .58) (PMC4974867). Hoffman 2012 reported a 12.7% squat-strength increase on phosphatidic acid versus 9.3% on placebo and 5.1% versus 3.3% for bench press, in 16 completers (PMC3506449)

Regulatory position: No EU health claim; no phosphatidic acid entry exists in the EU Register. The nearest entries are for phosphatidyl choline, all Non-authorised (entries 710, 1631, 1630, 709, EFSA 2010;8(10):1741), as is lecithin for memory and concentration (entry 1983) (EU Register)

Sources: PubMed 34706625 PMC3506449 PMC4974867 EU Register

Tongkat ali (Eurycoma longifolia)

It raises testosterone in the blood of adult men.

Claim as graded “Raises serum total testosterone in adult men”

Evidence tier B · provenance grade P3 · direction Contested · retail band 4, Experts disagree. Derived by rule 2 - direction Contested. Group: Ergogenic supplements. Scope qualifier: null in healthy eugonadal men

Best available evidence: PRISMA systematic review and random-effects meta-analysis of RCTs (Leisegang 2022, Nutrients, PMC9415500)

Sample and design: 521 articles identified, 9 studies in the systematic review, 5 RCTs in the meta-analysis; the review reports n = 232 in the pooled testosterone analysis, while its Table 2 sums to 267 (135 E. longifolia / 132 control) across 7 comparisons, because two trials contributed two arms each. Subgroups: healthy men n = 88 (45/43), hypogonadal men n = 179 (90/89). Doses administered ranged from 100 mg/day to 600 mg/day for 2 weeks to 6 months, and 1.7 mg/kg for 3 days to 5 weeks (PMC9415500)

Effect as reported: Overall SMD = 1.352, 95% CI 0.565 to 2.138, p = 0.001, but with severe instability: Q = 47.1472, DF = 6, p < 0.0001, I² = 87.27% (95% CI 76.06 to 93.24), and "a low p value (0.0243) was noticed with Begg's test, indicating publication bias". Split by population, the marketed claim collapses: normal healthy men SMD = 0.760, 95% CI −0.540 to 2.060, p = 0.249 (3 comparisons, n = 88) versus hypogonadal men SMD = 1.861, 95% CI 0.719 to 3.002, p = 0.002 (n = 179), the latter also showing significant publication bias on Egger's and Begg's tests. Individual SMDs ranged from −0.388 (95% CI −1.022 to 0.247) to 3.783 (95% CI 2.274 to 5.292) (PMC9415500)

Regulatory position: No EU health claim; Eurycoma longifolia does not appear in the EU Register (EU Register). A peer-reviewed regulatory review states "In the EU, E. longifolia is an unauthorized novel food" and that "the EFSA Panel concluded that E. longifolia has the potential to induce DNA damage, and its safety has not been established under any condition of use" (Food Reviews International 2023)

Safety signal: NIH LiverTox likelihood score: D (possible rare cause of clinically apparent liver injury). LiverTox documents a 47-year-old man who took a tongkat ali bodybuilding supplement for 2 weeks and developed hepatocellular injury with jaundice (R = 7.5, severity 3+, hospitalised): ALT rose from 470 to 875 U/L, alkaline phosphatase 192 to 238 U/L and total bilirubin 7.5 to 14.3 mg/dL against upper limits of 40 U/L, 125 U/L and 1.2 mg/dL, with recovery only partially documented (LiverTox, NBK609015). The meta-analysis notes one included trial reported gastrointestinal symptoms and itching (PMC9415500)

Sources: PMC9415500 LiverTox NBK609015 Food Reviews International 2023 EU Register

Testosterone, supraphysiologic doses (testosterone enanthate)

At doses above the natural range, it adds muscle and maximum strength in healthy men with normal testosterone, whether or not they lift weights.

Claim as graded “Supraphysiologic testosterone raises fat-free mass and maximal strength in healthy, non-hypogonadal men, with and without resistance training”

Evidence tier C · provenance grade P3 · direction Supported · retail band 2, Moderate evidence. Derived by rule 8 - tier C with provenance P3+. Group: Hormones and off-label pharmaceuticals.

Best available evidence: Single landmark double-blind randomised controlled trial: Bhasin et al., N Engl J Med 1996;335(1):1–7. Adjacent pooled evidence in hypogonadal/mixed populations: Corona et al., Eur J Endocrinol 2016;174:R99–R116

Sample and design: 43 normal men, 4 arms, 10 weeks, no pooling (NEJM 1996). Adjacent meta-analysis: 59 trials, 3029 treated / 2049 control, 40 studies reported lean mass (EJE 2016)

Effect as reported: No exercise, testosterone vs placebo: bench-press strength +9 ± 4 kg vs −1 ± 1 kg (P<0.05); squat strength +16 ± 4 kg vs +3 ± 1 kg (P<0.05); triceps area +424 ± 104 vs −81 ± 109 mm²; quadriceps area +607 ± 123 vs −131 ± 111 mm². Testosterone plus exercise: fat-free mass +6.1 ± 0.6 kg, bench press +22 ± 2 kg, squat +38 ± 4 kg. No 95% CIs were reported (NEJM 1996). Pooled lean-mass SMD in the adjacent meta-analysis, DEXA-derived trials: 0.55 (0.39–0.72), P<0.0001 (EJE 2016)

Regulatory position: Testosterone is a prescription-only licensed medicine in the EU/UK; the authorised indication is "Testosterone replacement therapy for male hypogonadism when testosterone deficiency has been confirmed by clinical features and biochemical tests", legal status "Prescription only medicine" (Nebido SmPC, emc). Use at supraphysiologic doses in healthy men falls outside that indication. Testosterone is listed by name in WADA 2026 Prohibited List S1.1 Anabolic Androgenic Steroids, prohibited at all times, non-Specified Substance (WADA 2026 Prohibited List)

Safety signal: Yes. Cross-sectional cohort of 86 long-term anabolic-androgenic steroid users vs 54 non-users: LVEF 52±11% vs 63±8% (P<0.001), E´ 9.3±2.4 vs 11.1±2.0 cm/s (P<0.001), coronary plaque volume 3 [0,174] vs 0 [0,69] mL³ (P=0.012); plaque rank rose 0.60 SD units (0.16–1.03) per 10 years of cumulative use (P=0.008) (Baggish et al., Circulation 2017;135:1991–2002). Danish register cohort, 1189 sanctioned AAS users vs 59,450 matched controls, mean 11 years follow-up: all-cause mortality HR 2.81 (1.98–3.99), P<.001; unnatural death HR 3.64 (2.22–5.96); natural death HR 2.24 (1.36–3.70) (Windfeld-Mathiasen et al., JAMA 2024;331(14):1229–1230)

Sources: Bhasin et al., N Engl J Med 1996;335(1):1–7 Corona et al., Eur J Endocrinol 2016;174:R99–R116 Nebido SmPC, emc WADA 2026 Prohibited List Baggish et al., Circulation 2017;135:1991–2002 Windfeld-Mathiasen et al., JAMA 2024;331(14):1229–1230

Nandrolone decanoate (19-nortestosterone decanoate)

It adds muscle and other non-fat tissue in adults aged 18 and over across a range of medical conditions.

Claim as graded “Nandrolone decanoate increases lean soft tissue mass in adults aged ≥18 across clinical populations”

Evidence tier B · provenance grade P3 · direction Supported · retail band 1, Strong evidence. Derived by rule 6 - tier A/B with provenance P3+. Group: Hormones and off-label pharmaceuticals.

Best available evidence: Systematic review and meta-analysis of RCTs vs placebo: Prokopidis et al., J Cachexia Sarcopenia Muscle 2026;17(2):e70276

Sample and design: 20 RCTs, adults ≥18, random-effects, RoB2, search to April 2025. Total participant count not stated on either fetched page (PubMed record; SCWD society abstract) — recorded as `n.a.`

Effect as reported: Lean soft tissue MD +1.59 kg (95% CI 1.06–2.13), p<0.01, I²=0%, k=11. Fat mass SMD −0.04, p=0.65, k=11 (no effect). Handgrip strength SMD 0.39, p=0.10, k=12 (no effect). Knee extension k=1, p=0.99 (no effect). GRADE: low certainty. No publication bias detected; most studies low risk of bias (J Cachexia Sarcopenia Muscle 2026)

Regulatory position: Licensed medicine in the EU by national procedure, not for muscle-building. Deca-Durabolin 50 mg/ml (nandrolone decanoate), Dutch marketing authorisation RVG 00126, first granted 20 May 1990: section 4.1 indications are "Behandeling van ernstige, d.w.z. klinisch manifeste, laat postmenopauzale osteoporose" (severe, clinically manifest late post-menopausal osteoporosis), "Anemie bij chronische nierinsufficiëntie" (anaemia of chronic renal insufficiency) and "Behandeling van aplastische anemie" (aplastic anaemia) (Dutch Medicines Evaluation Board SmPC). Lean-mass augmentation is not an authorised indication. Listed as "Nandrolone (19-nortestosterone)" in WADA 2026 Prohibited List S1.1, prohibited at all times, non-Specified Substance (WADA 2026)

Safety signal: Yes, class-level. The meta-analysis itself states nandrolone decanoate has "known safety concerns" but reports no numerical safety outcomes (J Cachexia Sarcopenia Muscle 2026). Class-level androgenic-anabolic steroid harms are documented above at H-01: 2.8-fold all-cause mortality in a 1189-user register cohort (JAMA 2024) and reduced LV systolic function plus accelerated coronary atherosclerosis in long-term users (Circulation 2017)

Sources: Prokopidis et al., J Cachexia Sarcopenia Muscle 2026;17(2):e70276 SCWD society abstract Dutch Medicines Evaluation Board SmPC WADA 2026 JAMA 2024 Circulation 2017

Oxandrolone (17α-methyl-2-oxa-dihydrotestosterone)

It adds muscle while recovering from severe burns — the group it was actually studied in.

Claim as graded “Oxandrolone increases lean body mass during the recovery phase in patients with severe burns — the population in which it was actually studied”

Evidence tier C · provenance grade P2 · direction Contested · retail band 4, Experts disagree. Derived by rule 2 - direction Contested. Group: Hormones and off-label pharmaceuticals.

Best available evidence: Systematic review and meta-analysis of RCTs: Front Med (Lausanne) 2025;12:1485474

Sample and design: 19 studies, 2,006 participants (779 oxandrolone, 1,227 standard care or placebo). Number of RCTs contributing to each individual outcome not stated (Front Med 2025)

Effect as reported: Lean body mass in recovery phase SMD 1.30, 95% CI −0.47 to 3.24, reported p<0.001, I²=95.0%. Weight gain in recovery SMD 0.58, 95% CI −1.21 to 2.38, p<0.001, I²=95.1%. Length of stay SMD −0.55 (−1.32 to 0.22), p<0.001, I²=97.3%. Donor-site healing time SMD −1.48 (−2.18 to −0.77), p=0.116. Mortality RR 1.04 (0.47–2.32) — the review states oxandrolone "did not significantly reduce mortality". Side effects RR 1.13 (0.68–1.87), p=0.174 (Front Med 2025). Note the internal inconsistency: the confidence intervals for the lean-body-mass, weight-gain and length-of-stay estimates all cross the null while the same table reports p<0.001 for each.

Regulatory position: Licensed medicine, but its approval has been withdrawn in the US and the approved indication was never muscle-building in healthy people. The FDA-approved indication was: "Oxandrin is indicated as adjunctive therapy to promote weight gain after weight loss following extensive surgery, chronic infections, or severe trauma, and in some patients who without definite pathophysiologic reasons fail to gain or to maintain normal weight, to offset the protein catabolism associated with prolonged administration of corticosteroids, and for the relief of the bone pain frequently accompanying osteoporosis" (Oxandrin label, DailyMed/FDA). FDA withdrew approval of NDA 013718 effective 28 June 2023 and determined the product "were withdrawn for reasons of safety or effectiveness", citing an advisory committee that "unanimously concluded that there was no evidence of efficacy for oxandrolone" and an "unfavorable" benefit-risk profile (Federal Register, 13 Sept 2023, 88 FR). No EU-authorised oxandrolone medicine was identified in this session (searched EMA medicine listings and the EU Union Register) — EU status recorded as `n.a.`. Listed as "Oxandrolone" in WADA 2026 Prohibited List S1.1, prohibited at all times (WADA 2026)

Safety signal: Yes — boxed warning. The label carries a boxed warning for peliosis hepatis ("sometimes ... associated with liver failure ... often not recognized until life-threatening liver failure or intra-abdominal hemorrhage develops"), liver cell tumors ("fatal malignant tumors have been reported"), and blood lipid changes including decreased HDL "known to be associated with increased risk of atherosclerosis" (DailyMed/FDA label). FDA additionally cited cholestatic hepatitis, hypercalcaemia in breast cancer, and increased risk of prostatic hypertrophy and prostatic carcinoma in geriatric patients (Federal Register 2023). Warfarin interaction study in 15 healthy subjects: S-warfarin half-life rose from 26 to 48 h and the warfarin dose needed to hold INR 1.5 fell 5.5-fold; microscopic haematuria in 9/15 (DailyMed/FDA label)

Sources: Front Med (Lausanne) 2025;12:1485474 Oxandrin label, DailyMed/FDA Federal Register, 13 Sept 2023, 88 FR WADA 2026

Dehydroepiandrosterone (DHEA, prasterone)

Taken by mouth, it raises testosterone in the blood of adults.

Claim as graded “Oral DHEA supplementation raises serum total testosterone concentration in adults”

Evidence tier B · provenance grade P3 · direction Supported · retail band 1, Strong evidence. Derived by rule 6 - tier A/B with provenance P3+. Group: Hormones and off-label pharmaceuticals.

Best available evidence: Two independent meta-analyses of RCTs: Exp Gerontol 2020;141:111110 and Diabetol Metab Syndr 2025;17:258

Sample and design: Primary: 42 publications, 55 arms; total participant count not stated on the fetched page → `n.a.` (Exp Gerontol 2020). Corroborating: 21 studies, 20 trial arms, 1,084 participants (519 DHEA / 565 control), postmenopausal women only (Diabetol Metab Syndr 2025)

Effect as reported: Overall WMD +28.02 ng/dL (95% CI 21.44–34.60), p=0.00. Subgroups (no CIs reported): women +30.98, men +21.36, >50 mg/d +57.96, ≤50 mg/d +19.43, healthy participants +52.17 ng/dL (Exp Gerontol 2020). Independent replication in postmenopausal women: WMD +24.31 ng/dL (15.22–33.40), p≤0.001, I²=99% (Diabetol Metab Syndr 2025)

Regulatory position: Prasterone (DHEA) is a prescription-only licensed medicine in the EU, but only as a vaginal pessary for a gynaecological indication. Intrarosa 6.5 mg pessary, EMEA/H/C/004138, "granted a marketing authorisation valid throughout the European Union ... on 8 January 2018", indicated "for the treatment of vulvar and vaginal atrophy in postmenopausal women having moderate to severe symptoms"; "The medicine can only be obtained with a prescription" (EMA, Intrarosa EPAR). No EU marketing authorisation for oral DHEA, and no authorised EU health claim for DHEA, was confirmed from a fetched page in this session — recorded as `n.a.` rather than asserted

Safety signal: `n.a.` — neither fetched meta-analysis reported safety findings or adverse events; the 42-publication review states only that "more study needed on pregnant women and miscarriage" (Exp Gerontol 2020). Absence of a reported signal in these two sources is not evidence of safety

Sources: Exp Gerontol 2020;141:111110 Diabetol Metab Syndr 2025;17:258 EMA, Intrarosa EPAR

Anastrozole, used off-label in men (aromatase inhibitor)

It raises testosterone in the blood of adult men whose levels are low because of weight or age.

Claim as graded “Anastrozole raises serum total testosterone in adult men with obesity- or ageing-related hypogonadism”

Evidence tier B · provenance grade P3 · direction Supported · retail band 1, Strong evidence. Derived by rule 6 - tier A/B with provenance P3+. Group: Hormones and off-label pharmaceuticals.

Best available evidence: Systematic review and meta-analysis, PROSPERO CRD12022296489: Dutta et al., 2024

Sample and design: 3 double-blind placebo-controlled RCTs, 118 participants at 3 months (Burnett-Bowie 2009 n=69; Colleluori 2020 n=23; Leder 2004/2005 n=26). Plus 3 uncontrolled studies, 52 participants, in the narrative review (Dutta et al.)

Effect as reported: Total testosterone at 3 months MD +7.08 nmol/L (95% CI 5.92–8.24), p<0.01, I²=0%, 3 RCTs, n=118, GRADE certainty High (control mean 10.43 nmol/L). At 6 months +6.61 nmol/L (5.30–7.93), 2 RCTs, n=92. At 12 months +5.20 nmol/L (3.78–6.62), 1 RCT, n=69. In the trials, anastrozole 1 mg/day was administered for 12 weeks, 6 months or 1 year (Dutta et al.)

Regulatory position: Prescription-only licensed medicine in the EU — but not for men. Arimidex 1 mg film-coated tablets, section 4.1: indicated for "Treatment of hormone receptor-positive advanced breast cancer in postmenopausal women", "Adjuvant treatment of hormone receptor-positive early invasive breast cancer in postmenopausal women", and adjuvant treatment after 2–3 years of tamoxifen; "Medicinal product subject to medical prescription" (EMA, Arimidex Article 30 referral Annex III). Use in men to raise testosterone is off-label. Listed as "Anastrozole" in WADA 2026 Prohibited List S4.1 Aromatase Inhibitors, prohibited at all times, Specified Substance (WADA 2026)

Safety signal: Yes, with imprecision. Pooled severe adverse events RR 2.48 (95% CI 0.42–14.66), p=0.32 — a point estimate 2.5× control that the 118-participant evidence base cannot exclude or confirm. The named severe adverse events occurring in the anastrozole arms were pancreatic carcinoma, hepatitis A, pulmonary embolism and embolic stroke; a further trial reported suicidal ideation (Dutta et al.). In a separate randomised comparison, one pulmonary embolism occurred in the anastrozole arm in a patient with prior DVT (Hohl et al., 2025)

Sources: Dutta et al., 2024 EMA, Arimidex Article 30 referral Annex III WADA 2026 Hohl et al., 2025

Enclomiphene / clomifene citrate, used off-label in men (SERMs)

It raises testosterone in adult men whose levels are low, and keeps sperm counts up better than testosterone gel does.

Claim as graded “SERM therapy raises serum total testosterone in adult men with total testosterone ≤300 ng/dL, and preserves sperm concentration relative to testosterone gel”

Evidence tier B · provenance grade P3 · direction Supported · retail band 1, Strong evidence. Derived by rule 6 - tier A/B with provenance P3+. Group: Hormones and off-label pharmaceuticals.

Best available evidence: Systematic review and meta-analysis of RCTs, PROSPERO CRD42024536930: Hohl et al., 2025. Supporting phase III programme: Wiehle et al., studies ZA-304 and ZA-305

Sample and design: 10 RCTs, 819 participants (374 SERM, 133 testosterone gel, 94 hCG, 13 anastrozole, 205 placebo); follow-up 2–30 weeks; mean baseline total testosterone 67–303 ng/dL (Hohl et al.). Phase III: two parallel randomised, double-blind, double-dummy, placebo-controlled 16-week multicentre studies in overweight men aged 18–60 with secondary hypogonadism; participant numbers not stated on the fetched record → `n.a.` (PubMed)

Effect as reported: Total testosterone, SERM vs placebo MD +273.76 ng/dL (95% CI 191.87–355.66), p<0.01, I²=89%, 5 studies, n=422. SERM vs testosterone gel +5.41 ng/dL (−43.44 to 54.27), p=0.83 — no difference. Sperm concentration, SERM vs testosterone gel +70.40 million/mL (41.62–99.18), p<0.01; change from baseline +55.18 million/mL (37.42–72.93), p<0.01; rate of men below 15 million/mL RR 0.10 (0.04–0.23). Sperm concentration, SERM vs placebo: +7.50 million/mL (−20.01 to 35.02), p=0.59 — no fertility benefit over placebo (Hohl et al.). Phase III reported directionally that enclomiphene raised LH, FSH and total testosterone and maintained sperm concentration in the normal range while testosterone gel produced "marked reduction in spermatogenesis", but no numeric effect sizes were reported on the fetched record (Wiehle et al.)

Regulatory position: Clomifene is a prescription-only licensed medicine in the EU/UK, licensed only for women. Clomifene 50 mg Tablets, PL29831/0037, section 4.1: "indicated for the treatment of ovulatory failure in women desiring pregnancy ... indicated only for patients in whom ovulatory dysfunction is demonstrated"; legal status "Prescription only medicine" (Clomifene 50mg SmPC, emc). Use in men is off-label. Enclomiphene: no EU-authorised enclomiphene medicine was identified in this session (searched EMA medicines listings and the EU Union Register); recorded as `n.a.` rather than asserted unlicensed. "Clomifene" is listed in WADA 2026 Prohibited List S4.2, anti-estrogenic substances and SERMs, prohibited at all times, Specified Substance (WADA 2026)

Safety signal: Yes, but underpowered. The review states its safety analysis "remained underpowered". Individual-trial signals: two deaths, one from stroke in the enclomiphene citrate group in a participant with multiple pre-existing risk factors; discontinuations for elevated haematocrit/haemoglobin and raised PSA in the 25 mg enclomiphene arm; a significant within-normal-range PSA rise on clomifene from 0.62±0.41 to 0.76±0.48 ng/mL, p=0.023 (Hohl et al.). An earlier review of 19 studies and 1,642 men reported side effects in under 10% and no serious adverse events (Huijben et al., Andrology 2022;10:451–469)

Sources: Hohl et al., 2025 Wiehle et al., studies ZA-304 and ZA-305 Clomifene 50mg SmPC, emc WADA 2026 Huijben et al., Andrology 2022;10:451–469

Human chorionic gonadotrophin (hCG) administered alongside testosterone therapy

Taken at a low dose alongside testosterone, it keeps the testicles working — both their own testosterone and sperm production.

Claim as graded “Concomitant low-dose hCG preserves testicular function — intratesticular testosterone and spermatogenesis — in men receiving exogenous testosterone”

Evidence tier D · provenance grade P2 · direction Insufficient · retail band 6, Too little evidence yet. Derived by rule 3 - direction Insufficient. Group: Hormones and off-label pharmaceuticals.

Best available evidence: No pooled evidence exists. Strongest single source is a small randomised dose-ranging trial using a surrogate endpoint: Coviello et al., J Clin Endocrinol Metab 2005;90(5):2595–2602. The only source measuring semen parameters is a retrospective series (Hsieh et al. 2013, as characterised in Desai et al., Ther Adv Urol 2022)

Sample and design: 29 men, normal reproductive physiology, 3 weeks, randomised to testosterone enanthate 200 mg weekly plus saline placebo or plus hCG 125, 250 or 500 IU every other day; intratesticular testosterone sampled by percutaneous fine-needle aspiration (Coviello et al.). Semen-parameter evidence: 26 men, retrospective series (Desai et al. 2022)

Effect as reported: Testosterone plus placebo suppressed intratesticular testosterone by 94%, from 1234 to 72 nmol/L; LH fell to 5% and FSH to 3% of baseline. Intratesticular testosterone rose linearly with hCG dose (P<0.001): post-treatment values were 25% below baseline (125 IU), 7% below baseline (250 IU) and 26% above baseline (500 IU). No 95% CIs reported. Sperm and semen parameters were not measured (Coviello et al.). The 26-man retrospective series reported "no change in semen parameters before treatment versus 1 year follow up", with no numeric values given (Desai et al. 2022)

Regulatory position: Prescription-only licensed medicine in the EU, for indications that do not include co-administration with testosterone therapy. Pregnyl 5000 IU, Irish MA PA0964/005/002, section 4.1 in the male: "Hypogonadotropic hypogonadism", "Delayed puberty associated with insufficient gonadotropic pituitary function", "Cryptorchidism, not due to anatomical obstruction", "Preoperative preparation of ectopic testes". The same SmPC states that during hCG treatment for hypogonadotropic hypogonadism "testosterone replacement therapy should be suspended" — the opposite of the graded claim (HPRA, Pregnyl 5000 IU SmPC). "chorionic gonadotrophin (CG)" is listed in WADA 2026 Prohibited List S2.2.1, testosterone-stimulating peptides in males, prohibited at all times (WADA 2026)

Safety signal: Not established either way from the fetched sources. The relevant documented harm is the harm the intervention is meant to mitigate: exogenous testosterone alone suppressed intratesticular testosterone by 94% within 3 weeks (Coviello et al.), and testosterone-induced azoospermia can occur within 10 weeks with up to 10% of men remaining azoospermic after cessation (Indications for hCG review, citing WHO Task Force 1990)

Sources: Coviello et al., J Clin Endocrinol Metab 2005;90(5):2595–2602 Hsieh et al. 2013, as characterised in Desai et al., Ther Adv Urol 2022 HPRA, Pregnyl 5000 IU SmPC WADA 2026 Indications for hCG review, citing WHO Task Force 1990

Semaglutide 2.4 mg once weekly (GLP-1 receptor agonist)

Alongside eating less and moving more, it takes weight off adults with obesity, or overweight plus a weight-related health problem, who do not have diabetes.

Claim as graded “Semaglutide 2.4 mg once weekly, as an adjunct to reduced-calorie diet and increased physical activity, reduces body weight in adults with obesity or overweight plus a weight-related comorbidity and without diabetes”

Evidence tier A · provenance grade P4 · direction Supported · retail band 1, Strong evidence. Derived by rule 6 - tier A/B with provenance P3+. Group: Hormones and off-label pharmaceuticals.

Best available evidence: Phase III double-blind RCT STEP 1, Wilding et al., N Engl J Med 2021, independently assessed within a full regulatory dossier by the EMA (Wegovy EPAR medicine overview)

Sample and design: 1,961 randomised (1,306 semaglutide / 655 placebo), 2:1, 68 weeks, 129 sites in 16 countries, participants without diabetes (NEJM 2021). EMA assessed five studies totalling 1,961 + 611 + 902 + 1,210 adults and 201 adolescents (EMA)

Effect as reported: Mean percentage change in body weight to week 68: −14.9% vs −2.4%; estimated treatment difference −12.4 percentage points (95% CI −13.4 to −11.5), P<0.001. Absolute weight −15.3 kg vs −2.6 kg; difference −12.7 kg (−13.7 to −11.7). ≥5% loss 86.4% vs 31.5% (P<0.001); ≥10% 69.1% vs 12.0%; ≥15% 50.5% vs 4.9%; ≥20% 32.0% vs 1.7% (NEJM 2021). EMA's independent restatement: "People treated with Wegovy lost on average 15% of their body weight, compared with a 2% loss in people who had placebo" (EMA)

Regulatory position: Licensed medicine in the EU for exactly this indication, prescription-only. "Wegovy received a marketing authorisation valid throughout the EU on 6 January 2022" (EMEA/H/C/005422). Indication: used with diet and physical activity to help adults lose weight and keep it under control, in adults with BMI ≥30 kg/m² (obesity) or BMI ≥27 to <30 kg/m² (overweight) with weight-related health problems, and in adolescents from 12 years with BMI ≥95th percentile weighing over 60 kg (EMA, Wegovy EPAR)

Safety signal: Yes, quantified and regulator-assessed. Serious adverse events 9.8% vs 6.4%; discontinuation for adverse events 7.0% vs 3.1%; gastrointestinal disorders 74.2% vs 47.9%; gallbladder-related disorders 2.6% vs 1.2%; cholelithiasis 1.8% vs 0.6%; acute pancreatitis 3 participants (0.2%) vs 0; nausea 44.2% vs 17.4%, vomiting 24.8% vs 6.6% (NEJM 2021). EMA lists headache, nausea, vomiting, diarrhoea, constipation and abdominal pain as affecting more than 1 in 10, and considers the side effects "manageable" (EMA)

Sources: STEP 1, Wilding et al., N Engl J Med 2021 Wegovy EPAR medicine overview

Tirzepatide once weekly (dual GIP/GLP-1 receptor agonist)

Alongside eating less and moving more, it takes weight off adults with obesity, or overweight plus at least one weight-related health problem.

Claim as graded “Tirzepatide once weekly, as an adjunct to reduced-calorie diet and increased physical activity, reduces body weight in adults with obesity, or overweight plus at least one weight-related comorbidity”

Evidence tier A · provenance grade P4 · direction Supported · retail band 1, Strong evidence. Derived by rule 6 - tier A/B with provenance P3+. Group: Hormones and off-label pharmaceuticals.

Best available evidence: Phase III double-blind RCT SURMOUNT-1, Jastreboff et al., N Engl J Med 2022;387:205–216, independently assessed by the EMA and reproduced in section 5.1 of the authorised product information (Mounjaro SmPC)

Sample and design: 2,539 randomised 1:1:1:1 to tirzepatide 5 mg (n=630), 10 mg (n=636), 15 mg (n=630) or placebo (n=643); 72 weeks; 119 sites in nine countries; December 2019 to April 2022; 86.0% completed (NEJM 2022)

Effect as reported: Treatment-regimen estimand, mean percentage weight change at week 72: 5 mg −15.0% (−15.9 to −14.2), difference vs placebo −11.9 pp (−13.4 to −10.4); 10 mg −19.5% (−20.4 to −18.5), difference −16.4 pp (−17.9 to −14.8); 15 mg −20.9% (−21.8 to −19.9), difference −17.8 pp (−19.3 to −16.3); placebo −3.1% (−4.3 to −1.9); P<0.001 for all comparisons. ≥20% weight reduction: 30.0% / 50.1% / 56.7% vs 3.1% (NEJM 2022). The EMA-authorised product information reproduces the efficacy estimand: −13.5 pp (−14.6 to −12.5), −18.9 pp (−20.0 to −17.8) and −20.1 pp (−21.2 to −19.0) vs placebo, p<0.001 adjusted for multiplicity (Mounjaro SmPC)

Regulatory position: Licensed medicine in the EU for exactly this indication, prescription-only. "Mounjaro received a marketing authorisation valid throughout the EU on 15 September 2022" (EMEA/H/C/005620); weight-management indication: "as an adjunct to a reduced-calorie diet and increased physical activity for weight management, including weight loss and weight maintenance, in adults with an initial Body Mass Index (BMI) of ≥ 30 kg/m2 (obesity) or ≥ 27 kg/m2 to < 30 kg/m2 (overweight) in the presence of at least one weight-related comorbid condition"; "can only be obtained with a prescription" (EMA, Mounjaro EPAR; Mounjaro SmPC)

Safety signal: Yes, quantified and regulator-assessed. Discontinuation for adverse events 4.3% / 7.1% / 6.2% vs 2.6% placebo; serious adverse events 6.3% / 6.9% / 5.1% vs 6.8%; deaths 4 / 2 / 1 vs 4; severe or serious gastrointestinal events 1.7% / 3.1% / 3.3% vs 1.1%; gallbladder disease 0.8% / 1.7% / 1.0% vs 0.8%; adjudicated pancreatitis 1 case in each arm (NEJM 2022). The SmPC carries warnings for acute pancreatitis ("Acute pancreatitis has been reported in patients treated with tirzepatide"), dehydration and acute renal failure secondary to gastrointestinal effects, diabetic retinopathy, and pulmonary aspiration under general anaesthesia or deep sedation (Mounjaro SmPC)

Sources: SURMOUNT-1, Jastreboff et al., N Engl J Med 2022;387:205–216 Mounjaro SmPC EMA, Mounjaro EPAR

Metformin (biguanide)

It extends healthy years of life, or cuts the chance of dying from any cause, in adults who do not have type 2 diabetes.

Claim as graded “Metformin extends healthspan or reduces all-cause mortality in adults who do not have type 2 diabetes”

Evidence tier C · provenance grade P3 · direction Not supported · retail band 5, Evidence says no. Derived by rule 1 - direction Not supported. Group: Hormones and off-label pharmaceuticals.

Best available evidence: Long-term randomised evidence from the only large placebo-controlled trial in a non-diabetic (prediabetic) population with adjudicated mortality follow-up: DPP/DPPOS, Diabetes Care 2021;44(12):2775–2782. The purpose-built ageing trial has not reported

Sample and design: 3,234 adults at high risk for type 2 diabetes randomised 1996–1999 (placebo n=1,082; metformin n=1,073; lifestyle n=1,079); median 21 years follow-up; 453 deaths. Metformin 850 mg twice daily was administered, masked during DPP and open-label during DPPOS (Diabetes Care 2021)

Effect as reported: All-cause mortality, metformin vs placebo: HR 0.99 (95% CI 0.79–1.25), P=0.95 (152 vs 143 deaths; 7.13 vs 6.59 per 1,000 person-years). Cancer mortality HR 1.04 (0.72–1.52); cardiovascular mortality HR 1.08 (0.70–1.66). Sensitivity analysis for out-of-study metformin use HR 0.97 (0.77–1.23); marginal structural model HR 0.78 (0.55–1.11) (Diabetes Care 2021). Observational pooled evidence points the other way but is confounded by indication and, as noted below, could not be transcribed reliably (Ageing Res Rev, Campbell et al.)

Regulatory position: Licensed medicine in the EU, prescription-only, for type 2 diabetes only. Glucophage 500/850 mg, PL 11648/0085-86, section 4.1: "Treatment of type 2 diabetes mellitus, particularly in overweight patients, when dietary management and exercise alone does not result in adequate glycaemic control"; legal status "Prescription only medicine" (Glucophage SmPC, emc). Use for healthspan or longevity in people without diabetes is off-label and outside any authorised indication. Metformin does not appear in WADA 2026 S1, S4.1 or S4.2 as listed on the fetched sections (WADA 2026)

Safety signal: No serious harm signal identified in the fetched trial source; DPP/DPPOS reported no excess all-cause, cancer or cardiovascular mortality (Diabetes Care 2021)

Sources: DPP/DPPOS, Diabetes Care 2021;44(12):2775–2782 Ageing Res Rev, Campbell et al. Glucophage SmPC, emc WADA 2026

Retatrutide (LY3437943)

It takes weight off adults with obesity, or overweight plus a weight-related condition, who do not have type 2 diabetes — more than a dummy injection does.

Claim as graded “Reduces body weight, measured on a calibrated scale, relative to placebo in adults with obesity or with overweight plus at least one weight-related condition and without type 2 diabetes”

Evidence tier C · provenance grade P4 · direction Supported · retail band 2, Moderate evidence. Derived by rule 8 - tier C with provenance P3+. Group: Hormones and off-label pharmaceuticals. Scope qualifier: body weight; no cardiovascular outcome trial published

Best available evidence: Two network meta-analyses include retatrutide, but both rest on the single published phase 2 trial: a Bayesian network meta-analysis of <strong>19 RCTs and 29,506 adults</strong> over at least 36 weeks (Obesity 2025;33:2046-2054) and a living systematic review and network meta-analysis of <strong>69 studies and 112,511 participants</strong>, searched to October 2025 (living SR/NMA). The only peer-reviewed primary trial is the <strong>phase 2 randomised, double-blind, placebo-controlled trial</strong> in NEJM (NEJM 2023; PubMed 37366315). The three phase 3 TRIUMPH trials exist only as company <strong>press releases</strong> with no peer review (TRIUMPH-1 press release; TRIUMPH-2 and -3 press release)

Sample and design: Phase 2 NCT04881760: <strong>338 adults enrolled</strong>, randomised 2:1:1:1:1:2:2 to placebo (70), 1 mg (69), 4 mg with 2 mg start (33), 4 mg with 4 mg start (34), 8 mg with 2 mg start (35), 8 mg with 4 mg start (35) or 12 mg (62); subcutaneous once weekly for 48 weeks; double-blind; primary endpoint percentage body-weight change at week 24 (NEJM 2023). Phase 3, press release only: TRIUMPH-1 NCT05929066, <strong>2,339 randomised 1:1:1:1</strong> to 4, 9 or 12 mg or placebo for 80 weeks, with a 104-week extension in 532 participants; TRIUMPH-2 NCT05929079, <strong>1,152</strong> adults with type 2 diabetes, 80 weeks; TRIUMPH-3 NCT05882045, <strong>1,949</strong> adults with BMI at least 35 and established cardiovascular disease (TRIUMPH-1; TRIUMPH-2 and -3)

Effect as reported: Phase 2 week 24, least-squares mean percentage body-weight change: 1 mg <strong>−7.2%</strong> (95% CI −8.5 to −5.9), combined 4 mg −11.8% (−13.3 to −10.2) and −13.9% (−15.9 to −11.9), combined 8 mg −16.7% (−18.4 to −15.1) and −17.9% (−19.7 to −16.1), 12 mg <strong>−17.5%</strong> (95% CI −18.8 to −16.1), placebo <strong>−1.6%</strong> (95% CI −2.7 to −0.5). <strong>No p-values are reported for the primary endpoint in the fetched NEJM page.</strong> Week 48: 1 mg −8.7%, combined 4 mg −17.1%, combined 8 mg −22.8%, 12 mg −24.2%, placebo −2.1% (NEJM; PubMed). Phase 3 TRIUMPH-1 week 80, efficacy estimand: −19.0% (4 mg), −25.9% (9 mg), <strong>−28.3%</strong> (12 mg) versus −2.2% placebo; treatment-regimen estimand −17.6%, −23.7%, −25.0% versus −3.9%. <strong>No confidence intervals and no p-values are stated in the press release.</strong> Reduction of at least 30% of body weight in 45.3% at 12 mg versus 0.5% on placebo; 104-week extension −30.3% (TRIUMPH-1). Pooled: retatrutide and other dual agonists −11.0 kg versus −9.0 kg for GLP-1 receptor agonists, odds ratio 54.6 for reduction of at least 15%; <strong>no confidence intervals or certainty ratings appear on the fetched abstract</strong> (Obesity 2025)

Regulatory position: <strong>Not approved anywhere.</strong> Drugs@FDA returns "No matches found!" for retatrutide (openFDA Drugs@FDA query). Retatrutide appears in the openFDA NDC directory in <strong>12 entries, every one with marketing category "BULK INGREDIENT" and product type "BULK INGREDIENT"</strong>, listed by ingredient manufacturers including Shanghai SynTheAll, Nanjing Chengong, Qingdao Biopeptek, Hybio and Harbin Jixianglong (openFDA NDC query). A bulk-ingredient NDC listing is a directory entry made by a substance supplier and <strong>is not a marketing authorisation, an approval, or any FDA finding about the substance</strong>. The EMA medicines register, generated 03/09/2026 and containing 2,741 rows, has <strong>zero entries</strong> for retatrutide (EMA medicines register). The sponsor states it "plans to submit" a Biologics License Application in the first quarter of 2027 (Lilly). Anti-doping status: <strong>n.a.</strong> — the WADA Prohibited List PDF was not fetched in this session

Safety signal: Phase 2, with denominators: any adverse event in <strong>277 of 337 (82%)</strong> treated participants overall, <strong>49 of 70 (70%)</strong> on placebo, <strong>57 of 62 (92%)</strong> at 12 mg and 33 of 35 (94%) at 8 mg with a 4 mg start. Discontinuation for an adverse event in <strong>27 of 337 (8%)</strong> overall, <strong>0 of 70 (0%)</strong> on placebo, <strong>10 of 62 (16%)</strong> at 12 mg and 5 of 35 (14%) at 8 mg with a 2 mg start; serious adverse events in 3 of 70 (4%) on placebo. Gastrointestinal events were dose-related and were partially mitigated by a lower starting dose; dose-dependent heart-rate increases peaked at week 24 (NEJM; PubMed). Phase 3 TRIUMPH-1, 12 mg versus placebo, <strong>percentages given without denominators</strong>: nausea 42.4% versus 14.8%, vomiting 25.3% versus 4.8%, dysesthesia 12.5% versus 0.9% (TRIUMPH-1). TRIUMPH-2 discontinuation for adverse events 3.8%, 11.6% and 7.7% versus 4.9% on placebo, <strong>denominators not stated</strong> (TRIUMPH-2 and -3). The Bayesian network meta-analysis records retatrutide as carrying the <strong>highest adverse-event risk</strong> of the agents compared (Obesity 2025)

Sources: Obesity 2025;33:2046-2054 living SR/NMA NEJM 2023 PubMed 37366315 TRIUMPH-1 press release TRIUMPH-2 and -3 press release openFDA Drugs@FDA query openFDA NDC query EMA medicines register

Orforglipron (LY3502970)

Taken as a daily pill, it takes weight off adults with obesity — more than a dummy pill does.

Claim as graded “Reduces body weight, measured on a calibrated scale, relative to placebo in adults with obesity, taken orally once daily”

Evidence tier B · provenance grade P4 · direction Supported · retail band 1, Strong evidence. Derived by rule 6 - tier A/B with provenance P3+. Group: Hormones and off-label pharmaceuticals. Scope qualifier: body weight; cardiovascular outcome data not located

Best available evidence: A <strong>living systematic review and network meta-analysis</strong> of 69 studies and 112,511 participants, searched to October 2025, includes orforglipron among the compared interventions (living SR/NMA), together with an <strong>independent regulatory assessment</strong>: FDA approval of the compound under application NDA220934 (openFDA Drugs@FDA; FDA press announcement). Two peer-reviewed phase 3 trials: ATTAIN-1 (NEJM, PubMed 40960239) and ATTAIN-2 (PubMed 41275875)

Sample and design: ATTAIN-1 NCT05869903: <strong>3,127 patients randomised 3:3:3:4</strong> to 6, 12 or 36 mg once daily or placebo, phase 3, multinational, double-blind, 72 weeks, adults with obesity without diabetes; primary endpoint percentage body-weight change to week 72 on the treatment-regimen estimand in the intention-to-treat population (PubMed 40960239); dose escalation began at 1 mg once daily with four-weekly steps (Lilly ATTAIN-1 press release). ATTAIN-2 NCT05872620: <strong>1,613 randomised</strong> (6 mg 329, 12 mg 332, 36 mg 322, placebo 630), 72 weeks, adults with obesity or overweight and type 2 diabetes (PubMed 41275875)

Effect as reported: ATTAIN-1 week 72, treatment-regimen estimand: 6 mg −7.5% (95% CI −8.2 to −6.8), 12 mg −8.4% (95% CI −9.1 to −7.7), <strong>36 mg −11.2%</strong> (95% CI −12.0 to −10.4) versus <strong>placebo −2.1%</strong> (95% CI −2.8 to −1.4), <strong>P&lt;0.001</strong> for all comparisons with placebo. At 36 mg, 54.6% had a reduction of at least 10%, 36.0% at least 15% and 18.4% at least 20%, versus 12.9%, 5.9% and 2.8% on placebo (PubMed 40960239). ATTAIN-2 week 72: −5.1% (95% CI −6.0 to −4.2), −7.0% (−7.8 to −6.2), <strong>−9.6%</strong> (−10.5 to −8.7) versus <strong>−2.5%</strong> (−3.0 to −1.9), all <strong>p&lt;0.0001</strong> (PubMed 41275875). The living network meta-analysis reports no numeric pooled estimate for orforglipron on the fetched page and no GRADE certainty rating (living SR/NMA)

Regulatory position: <strong>Approved in the United States.</strong> Drugs@FDA lists application NDA220934, submission ORIG-1, status "AP" dated 20260401, brand name FOUNDAYO (openFDA Drugs@FDA). FDA states verbatim that "Foundayo is approved for use in combination with a reduced-calorie diet and increased physical activity to reduce excess body weight and maintain weight reduction long term in adults with obesity or adults with overweight in the presence of at least one weight-related comorbid condition", and that the approval came 50 days after filing under the Commissioner's National Priority Voucher pilot, 294 days before the PDUFA date of January 20, 2027. The label carries a <strong>boxed warning for thyroid C-cell tumors</strong> (FDA). <strong>Not authorised in the EU</strong>: zero entries for orforglipron in the EMA medicines register generated 03/09/2026 (EMA register). Anti-doping status: <strong>n.a.</strong>

Safety signal: Discontinuation for adverse events in <strong>5.3% to 10.3%</strong> of orforglipron-treated patients versus <strong>2.7%</strong> on placebo in ATTAIN-1; <strong>denominators are not stated in the fetched abstract</strong> (PubMed 40960239). The sponsor's press release gives per-arm figures of 5.1%, 7.7% and 10.3% versus 2.6% on placebo, again <strong>without denominators</strong>, alongside nausea 28.9% / 35.9% / 33.7% versus 10.4%, constipation 21.7% / 29.8% / 25.4% versus 9.3%, vomiting 13.0% / 21.4% / 24.0% versus 3.5%; the release states that <strong>overall treatment-emergent adverse event totals, serious adverse event counts and their denominators are not provided</strong> (Lilly). ATTAIN-2: discontinuation for adverse events 6.1% to 9.9% versus 4.1% on placebo, denominators not stated, and <strong>10 deaths (6 on orforglipron, 4 on placebo)</strong> (PubMed 41275875). Regulator wording: FDA imposed a boxed warning for thyroid C-cell tumors (FDA)

Sources: living SR/NMA openFDA Drugs@FDA FDA press announcement NEJM, PubMed 40960239 PubMed 41275875 Lilly ATTAIN-1 press release EMA register

Cagrilintide (AM833), as monotherapy

On its own, it takes weight off adults without diabetes who have obesity, or overweight with high blood pressure or high cholesterol — more than a dummy injection does.

Claim as graded “Reduces body weight, measured on a calibrated scale, relative to placebo in adults without diabetes who have obesity or overweight with hypertension or dyslipidaemia”

Evidence tier C · provenance grade P4 · direction Supported · retail band 2, Moderate evidence. Derived by rule 8 - tier C with provenance P3+. Group: Hormones and off-label pharmaceuticals. Scope qualifier: body weight over 26 to 68 weeks

Best available evidence: <strong>None for this claim.</strong> No synthesis of cagrilintide monotherapy was located. The evidence is two randomised comparisons: the <strong>phase 2 dose-finding trial</strong> (Lancet 2021, Europe PMC record; PubMed 34798060) and the cagrilintide-alone arm of the <strong>phase 3a REDEFINE 1 trial</strong> (NEJM REDEFINE 1; REDEFINE 1 secondary report, PMC)

Sample and design: Phase 2 NCT03856047: <strong>706 participants randomised to cagrilintide 0.3, 0.6, 1.2, 2.4 or 4.5 mg</strong> (100–102 per dose group), <strong>99 to liraglutide 3.0 mg</strong> and <strong>101 to volume-matched placebo</strong> in six placebo groups, randomisation 6:1, at 57 sites in ten countries; subcutaneous self-injection once weekly for a 26-week treatment period including dose escalation of up to 6 weeks, with 6 weeks of follow-up; participants and investigators were masked to active versus pooled placebo but not to different active treatments; primary endpoint percentage body-weight change from baseline to week 26 (Europe PMC record of Lancet 2021). REDEFINE 1 NCT05567796 included a <strong>cagrilintide 2.4 mg monotherapy arm of 302 participants</strong> within a 3,417-participant, 68-week, phase 3a trial (NEJM REDEFINE 1)

Effect as reported: Phase 2 week 26, trial-product estimand: mean reductions of <strong>6.0% to 10.8% (6.4 to 11.5 kg) across 0.3 to 4.5 mg</strong> versus <strong>3.0% (3.3 kg) on placebo</strong>, estimated treatment difference range 3.0% to 7.8%, <strong>p&lt;0.001</strong>; cagrilintide 4.5 mg −10.8% (11.5 kg) versus liraglutide 3.0 mg −9.0% (9.6 kg), estimated treatment difference 1.8%, <strong>p=0.03</strong>. <strong>No 95% confidence intervals are reported in the fetched record.</strong> Results with the treatment-policy estimand were similar (Europe PMC record). REDEFINE 1 week 68: cagrilintide 2.4 mg <strong>−11.8%</strong> versus placebo −2.3% (REDEFINE 1 secondary report); <strong>no confidence interval or p-value for the cagrilintide arm was located in any fetched source</strong>

Regulatory position: <strong>Not approved anywhere as monotherapy.</strong> Drugs@FDA returns "No matches found!" for cagrilintide (openFDA Drugs@FDA). Three openFDA NDC directory entries exist, <strong>all with marketing category and product type "BULK INGREDIENT"</strong>, listed by ingredient suppliers; this is a supplier directory listing and <strong>not an approval</strong> (openFDA NDC). Zero entries in the EMA medicines register generated 03/09/2026 (EMA register). The only regulatory filing located concerns the co-formulation, not monotherapy (Novo Nordisk REDEFINE announcement). Anti-doping status: <strong>n.a.</strong>

Safety signal: Phase 2: <strong>permanent treatment discontinuation in 73 of 906 randomised participants (10%)</strong>, similar across treatment groups, mostly for adverse events (<strong>30 participants, 4%</strong>); 29 participants (4%) withdrew from the trial. Gastrointestinal adverse events in <strong>41% to 63% of cagrilintide 0.3–4.5 mg participants versus 32% on placebo</strong>, primarily nausea, <strong>20% to 47% versus 18%</strong>; administration-site reactions were also among the most frequent events. <strong>Arm-level denominators for the adverse-event percentages are not given in the fetched record</strong> (Europe PMC record). REDEFINE 1 reported adverse events by arm in a supplementary table that was not accessible; arm-specific totals for the cagrilintide monotherapy arm are <strong>n.a.</strong> — searched the NEJM record, the PMC secondary report and the sponsor's REDEFINE announcement (PMC secondary report)

Sources: Lancet 2021, Europe PMC record PubMed 34798060 NEJM REDEFINE 1 REDEFINE 1 secondary report, PMC openFDA Drugs@FDA openFDA NDC EMA register Novo Nordisk REDEFINE announcement

CagriSema (cagrilintide 2.4 mg co-formulated with semaglutide 2.4 mg)

Over 68 weeks it takes weight off adults with overweight or obesity who do not have type 2 diabetes — more than a dummy injection does.

Claim as graded “Reduces body weight, measured on a calibrated scale, relative to placebo at week 68 in adults with overweight or obesity and without type 2 diabetes”

Evidence tier C · provenance grade P4 · direction Supported · retail band 2, Moderate evidence. Derived by rule 8 - tier C with provenance P3+. Group: Hormones and off-label pharmaceuticals. Scope qualifier: body weight; cardiovascular outcome trial not reported

Best available evidence: <strong>No synthesis reporting a pooled estimate for this combination was located.</strong> The living systematic review and network meta-analysis lists semaglutide-cagrilintide among included interventions but the fetched record gives <strong>no numeric pooled estimate and no certainty rating</strong> for it (living SR/NMA). Primary evidence is the <strong>phase 3a REDEFINE 1 trial</strong> in NEJM (NEJM REDEFINE 1; REDEFINE 1 PDF; secondary report, PMC), with REDEFINE 2 reported only in a company release (Novo Nordisk)

Sample and design: REDEFINE 1 NCT05567796, phase 3a, 68 weeks: <strong>3,417 participants randomised 21:3:3:7</strong> to cagrilintide-semaglutide (2,108), semaglutide 2.4 mg alone (302), cagrilintide 2.4 mg alone (302) or placebo (705), plus lifestyle intervention in all groups; enrolment required a BMI of 30 or higher, or 27 or higher with at least one obesity-related complication; coprimary endpoints were relative body-weight change and a reduction of at least 5% at week 68, assessed on the treatment-policy estimand (NEJM REDEFINE 1; PMC). REDEFINE 2, <strong>1,206 participants with type 2 diabetes</strong>, is reported only in a company release (Novo Nordisk)

Effect as reported: Week 68, treatment-policy estimand: <strong>−20.4%</strong> with cagrilintide-semaglutide versus <strong>−3.0%</strong> with placebo, estimated difference <strong>−17.3 percentage points</strong> (95% CI −18.1 to −16.6), <strong>P&lt;0.001</strong>. Trial-product estimand −22.7% versus −2.3%, difference −20.4 percentage points (95% CI −21.1 to −19.7). Reductions of at least 5% in 91.9% versus 31.5%, at least 20% in 53.6% versus 1.9%, at least 30% in 19.3% versus 0.4%, <strong>P&lt;0.001 for all comparisons</strong> (REDEFINE 1 PDF; NEJM record). Comparator arms at week 68 on the trial-product estimand: semaglutide 2.4 mg −16.1%, cagrilintide 2.4 mg −11.8% (PMC). The result fell short of the sponsor's pre-announced 25% expectation, which is <strong>a material discrepancy between the trial outcome and prior guidance and is recorded here rather than omitted</strong> (REDEFINE 1 PDF)

Regulatory position: <strong>Not approved in the United States or the EU.</strong> A New Drug Application was submitted to FDA on 18 December 2025 and the compound is described as under review, not authorised (Novo Nordisk). The EMA medicines register generated 03/09/2026 contains <strong>no entry for cagrilintide or for any cagrilintide-semaglutide product</strong>; the only semaglutide entries are Wegovy (EMEA/H/C/005422, Authorised, indications Obesity and Overweight), Ozempic, Rybelsus, Kayshild and Kyinsu (EMA register). Anti-doping status: <strong>n.a.</strong>

Safety signal: Gastrointestinal adverse events, including nausea, vomiting, diarrhoea, constipation or abdominal pain, in <strong>79.6% of the cagrilintide-semaglutide group and 39.9% of the placebo group</strong>, described as mainly transient and mild-to-moderate; <strong>denominators for these percentages are not stated in the fetched abstract</strong>, though the arm sizes were 2,108 and 705 (NEJM record). Hypotension in <strong>38 participants (1.8%)</strong> on cagrilintide-semaglutide versus <strong>3 (0.4%)</strong> on placebo (PMC secondary report). Company release: nausea 55% versus 12.6%, vomiting 26.1% versus 4.1%, discontinuation for adverse events <strong>5.9% versus 3.5%</strong> in REDEFINE 1 and <strong>8.4% versus 3%</strong> in REDEFINE 2, <strong>all without denominators</strong> (Novo Nordisk). Overall serious adverse event counts by arm are <strong>n.a.</strong> — the trial reported them in a supplementary table that was not accessible; searched the NEJM record, the NEJM PDF mirror and the PMC secondary report

Sources: living SR/NMA NEJM REDEFINE 1 REDEFINE 1 PDF secondary report, PMC Novo Nordisk EMA register

Survodutide (BI 456906)

Over 76 weeks it takes weight off adults with obesity who do not have diabetes — more than a dummy injection does.

Claim as graded “Reduces body weight, measured on a calibrated scale, relative to placebo at week 76 in adults with obesity and without diabetes”

Evidence tier C · provenance grade P4 · direction Supported · retail band 2, Moderate evidence. Derived by rule 8 - tier C with provenance P3+. Group: Hormones and off-label pharmaceuticals. Scope qualifier: body weight; no cardiovascular outcome result located

Best available evidence: <strong>No synthesis specific to survodutide was located.</strong> A Bayesian network meta-analysis of 19 RCTs and 29,506 adults includes survodutide but pools it with other dual and triple agonists rather than estimating it separately, and reports <strong>no confidence intervals or certainty ratings on the fetched record</strong> (Obesity 2025). Primary evidence is the <strong>phase 3 SYNCHRONIZE-1 trial</strong> in NEJM (NEJM 2026;395:776-787, PubMed 42253238; Europe PMC record) and the <strong>phase 2 dose-finding trial</strong> in the Lancet (PubMed 38330987)

Sample and design: SYNCHRONIZE-1 NCT06066515, phase 3, 76 weeks: <strong>725 participants randomised 1:1:1</strong> to survodutide adjusted up to 3.6 mg (241) or 6.0 mg (242) subcutaneously once weekly, or placebo (242), with counselling for lifestyle modification; mean age 47.1 years, 294 participants (40.6%) men, mean BMI 37.9, mean body weight 108.8 kg; two primary endpoints, percentage body-weight change and a reduction of at least 5% from baseline to week 76, analysed on the treatment-regimen estimand (Europe PMC record). Phase 2 NCT04667377: <strong>387 enrolled and 386 treated</strong> (0.6 mg 77, 2.4 mg 78, 3.6 mg 77, 4.8 mg 77, placebo 77) for 46 weeks (PubMed 38330987)

Effect as reported: SYNCHRONIZE-1 week 76, treatment-regimen estimand: <strong>−12.2%</strong> (95% CI −13.6 to −10.8) at 3.6 mg and <strong>−13.0%</strong> (95% CI −14.4 to −11.6) at 6.0 mg versus <strong>−5.4%</strong> (95% CI −6.9 to −4.0) on placebo, with a reduction of at least 5% in 72.6%, 71.9% and 46.3% respectively, <strong>P&lt;0.001 for all comparisons with placebo</strong> (Europe PMC record; PubMed 42253238). Phase 2 week 46: −6.2% (95% CI −8.3 to −4.1), −12.5% (−14.5 to −10.5), −13.2% (−15.3 to −11.2), −14.9% (−16.9 to −13.0) versus placebo −2.8% (−4.9 to −0.7); <strong>no p-values appear on the fetched page</strong> (PubMed 38330987). The sponsor's release headlines a <strong>16.6% reduction versus 3.2% on placebo using the efficacy estimand</strong>, a larger figure than the pre-specified primary treatment-regimen estimand; <strong>the smaller primary-estimand figure of −13.0% is published here as the graded result and the headline figure is shown alongside</strong> (Boehringer Ingelheim press release)

Regulatory position: <strong>Not approved anywhere.</strong> Drugs@FDA returns "No matches found!" for survodutide and the openFDA NDC directory returns <strong>no records at all</strong>, so not even a bulk-ingredient listing exists (openFDA Drugs@FDA; openFDA NDC). Zero entries in the EMA medicines register generated 03/09/2026 (EMA register). The sponsor states verbatim: "Survodutide is an investigational agent and has not been approved for use; its efficacy and safety has not been established" (Boehringer Ingelheim). Anti-doping status: <strong>n.a.</strong>

Safety signal: SYNCHRONIZE-1: gastrointestinal symptoms, typically mild to moderate, in <strong>80.9% of the 3.6 mg group (n=241), 89.7% of the 6.0 mg group (n=242) and 47.9% of the placebo group (n=242)</strong>; <strong>no deaths were reported</strong>. Overall treatment-emergent adverse event totals, serious adverse event counts and discontinuation-for-adverse-event rates are <strong>n.a.</strong> — searched the NEJM full text, which returned a client error, the PubMed abstract and the Europe PMC record (Europe PMC record). Phase 2, with denominators: adverse events in <strong>281 of 309 (91%)</strong> survodutide participants versus <strong>58 of 77 (75%)</strong> on placebo, and <strong>only 233 of 386 (60.4%) completed treatment</strong>, a completion rate that materially limits interpretation of the 46-week estimates (PubMed 38330987)

Sources: Obesity 2025 NEJM 2026;395:776-787, PubMed 42253238 Europe PMC record PubMed 38330987 Boehringer Ingelheim press release openFDA Drugs@FDA openFDA NDC EMA register

Mazdutide (IBI362, LY3305677)

Over 32 weeks it takes weight off Chinese adults with obesity or overweight — more than a dummy injection does.

Claim as graded “Reduces body weight, measured on a calibrated scale, relative to placebo at week 32 in Chinese adults with obesity or overweight”

Evidence tier C · provenance grade P4 · direction Supported · retail band 2, Moderate evidence. Derived by rule 8 - tier C with provenance P3+. Group: Hormones and off-label pharmaceuticals. Scope qualifier: body weight; population confined to Chinese adults

Best available evidence: <strong>None for this claim.</strong> No synthesis of mazdutide was located. Primary evidence is the single <strong>phase 3 GLORY-1 trial</strong> published in NEJM (PubMed 40421736; mirrored copy of the NEJM article text), together with an <strong>independent regulatory assessment</strong> by China's NMPA (Innovent announcement; PubMed 41028652)

Sample and design: GLORY-1 NCT05607680, phase 3, double-blind, placebo-controlled, conducted in China: <strong>610 participants randomised 1:1:1</strong> to mazdutide 4 mg (203), mazdutide 6 mg (202) or placebo (205), subcutaneously once weekly for 48 weeks. The 4 mg arm received 2 mg during weeks 1–4 then 4 mg during weeks 5–48; the 6 mg arm received 2 mg during weeks 1–4, 4 mg during weeks 5–8 and 6 mg during weeks 9–48. Primary endpoint percentage body-weight change at week 32 (mirrored NEJM article text; PubMed 40421736)

Effect as reported: Week 32, primary endpoint: 4 mg <strong>−10.09%</strong> (95% CI −11.15 to −9.04), 6 mg <strong>−12.55%</strong> (95% CI −13.64 to −11.45), placebo <strong>+0.45%</strong> (95% CI −0.61 to 1.52), <strong>P&lt;0.001</strong> for both comparisons. Week 48: −11.00% (95% CI −12.27 to −9.73), −14.01% (95% CI −15.36 to −12.66), placebo +0.30%. Reduction of at least 5% at week 32 in 73.9% and 82.0% versus 10.5% on placebo (PubMed 40421736)

Regulatory position: <strong>Approved in China only.</strong> China's NMPA granted approval on 27 June 2025 for chronic weight management in adults with a BMI of at least 28, or at least 24 with a weight-related comorbidity, under the brand name Xinermei (Innovent); a type 2 diabetes indication was approved in September 2025 (PubMed 41028652). <strong>Not approved in the United States</strong>: Drugs@FDA returns "No matches found!" for mazdutide, and the single openFDA NDC record is marketing category and product type <strong>"BULK INGREDIENT"</strong>, which is a supplier directory listing and <strong>not an approval</strong> (openFDA Drugs@FDA; openFDA NDC). <strong>Not authorised in the EU</strong>: zero entries in the EMA medicines register generated 03/09/2026 (EMA register). Anti-doping status: <strong>n.a.</strong>

Safety signal: Complete denominators are available for this trial. Any adverse event in <strong>195 of 203 (96.1%)</strong> at 4 mg, <strong>196 of 202 (97.0%)</strong> at 6 mg and <strong>183 of 205 (89.3%)</strong> on placebo. Serious adverse events in <strong>12 of 203 (5.9%)</strong>, <strong>8 of 202 (4.0%)</strong> and <strong>13 of 205 (6.3%)</strong>. Discontinuation of mazdutide or placebo for an adverse event in <strong>3 of 203 (1.5%)</strong>, <strong>1 of 202 (0.5%)</strong> and <strong>2 of 205 (1.0%)</strong>. Nausea 66 of 203 (32.5%), 102 of 202 (50.5%) and 12 of 205 (5.9%); diarrhoea 71 of 203 (35.0%), 78 of 202 (38.6%) and 13 of 205 (6.3%); vomiting 53 of 203 (26.1%), 87 of 202 (43.1%) and 6 of 205 (2.9%); abdominal distention 13 of 203 (6.4%), 28 of 202 (13.9%) and 4 of 205 (2.0%). Events were mostly mild or moderate and occurred during dose escalation, with incidence rising at the higher dose (mirrored NEJM article text)

Sources: PubMed 40421736 mirrored copy of the NEJM article text Innovent announcement PubMed 41028652 openFDA Drugs@FDA openFDA NDC EMA register

Ecnoglutide (XW003)

Over 40 weeks it takes weight off Chinese adults with obesity or overweight who do not have diabetes — more than a dummy injection does.

Claim as graded “Reduces body weight, measured on a calibrated scale, relative to placebo at week 40 in Chinese adults with obesity or overweight and without diabetes”

Evidence tier C · provenance grade P4 · direction Supported · retail band 2, Moderate evidence. Derived by rule 8 - tier C with provenance P3+. Group: Hormones and off-label pharmaceuticals. Scope qualifier: body weight; population confined to Chinese adults

Best available evidence: <strong>None for this claim.</strong> No synthesis of ecnoglutide was located. Primary evidence is a single <strong>phase 3 randomised, double-blind, placebo-controlled trial</strong> in Lancet Diabetes and Endocrinology (PubMed 40555243; Europe PMC record), plus an <strong>independent regulatory assessment</strong> by China's NMPA (PubMed 42412371)

Sample and design: NCT05813795, phase 3, 36 centres in China, 40 weeks: <strong>664 participants randomised</strong> to ecnoglutide 1.2 mg (166), 1.8 mg (166), 2.4 mg (167) or placebo (165), in adults with obesity or overweight and without type 1 or type 2 diabetes; primary endpoint percentage body-weight change at week 40 (PubMed 40555243)

Effect as reported: Week 40, least-squares mean percentage change: <strong>−9.1% (SE 0.8)</strong> at 1.2 mg, <strong>−10.9% (SE 0.9)</strong> at 1.8 mg, <strong>−13.2% (SE 0.8)</strong> at 2.4 mg versus <strong>+0.1% (SE 0.8)</strong> on placebo; estimated treatment differences <strong>−9.2%</strong> (97% CI −11.0 to −7.5), <strong>−11.1%</strong> (97% CI −13.1 to −9.1) and <strong>−13.3%</strong> (97% CI −15.3 to −11.3), <strong>all p&lt;0.0001</strong>. Reduction of at least 5% in 77%, 84% and 87% versus 16% on placebo, treatment differences 60% (98% CI 50 to 71), 68% (58 to 78) and 70% (61 to 80), all p&lt;0.0001 (Europe PMC record)

Regulatory position: <strong>Approved in China only.</strong> NMPA approved ecnoglutide in January 2026 for glycaemic control in type 2 diabetes and in March 2026 for long-term weight management in adults with obesity, or overweight with at least one weight-related comorbidity (PubMed 42412371). <strong>Not approved in the United States</strong>: Drugs@FDA returns "No matches found!", and the single openFDA NDC record is marketing category and product type <strong>"BULK INGREDIENT"</strong>, a supplier directory listing that is <strong>not an approval</strong> (openFDA Drugs@FDA; openFDA NDC). <strong>Not authorised in the EU</strong>: zero entries in the EMA medicines register generated 03/09/2026 (EMA register). Anti-doping status: <strong>n.a.</strong>

Safety signal: Treatment-emergent adverse events in <strong>155 of 166 (93%)</strong> at 1.2 mg, <strong>154 of 166 (93%)</strong> at 1.8 mg, <strong>156 of 167 (93%)</strong> at 2.4 mg and <strong>139 of 165 (84%)</strong> on placebo; the most common were mild-to-moderate gastrointestinal events. <strong>Ten ecnoglutide participants discontinued treatment because of adverse events; the fetched record gives no arm split and no denominator for that count</strong>, and <strong>serious adverse event counts are n.a.</strong> — searched the PubMed abstract, the Europe PMC core record and the Lancet full text, which returned a client error (Europe PMC record)

Sources: PubMed 40555243 Europe PMC record PubMed 42412371 openFDA Drugs@FDA openFDA NDC EMA register

Petrelintide (ZP8396)

Over 28 weeks it takes weight off adults with overweight or obesity — more than a dummy injection does.

Claim as graded “Reduces body weight, measured on a calibrated scale, relative to placebo at week 28 in adults with overweight or obesity”

Evidence tier C · provenance grade P4 · direction Insufficient · retail band 6, Too little evidence yet. Derived by rule 3 - direction Insufficient. Group: Hormones and off-label pharmaceuticals. Scope qualifier: the doses administered were not disclosed and no peer-reviewed report exists

Best available evidence: <strong>None for this claim, and no peer-reviewed publication exists.</strong> The only evidence is a company <strong>press release and investor presentation</strong> reporting phase 2 topline results; no journal article was located (Zealand Pharma press release; ZUPREME-1 conference-call presentation)

Sample and design: ZUPREME-1 NCT06662539, phase 2, 33 sites in the United States, Poland and Romania: <strong>493 people</strong> randomised to five dose groups or placebo, with a primary endpoint of percentage body-weight change at week 28, maintenance to week 42 and follow-up to week 51. Arm sizes from the sponsor presentation: placebo 81, dose groups 79, 81, 83, 79 and 82, all petrelintide arms combined 404, highest-response dose arm 83. <strong>The numerical doses administered were not disclosed in either fetched source</strong>, so the exposure element of the claim cannot be stated (Zealand Pharma; presentation)

Effect as reported: Mean reduction of <strong>up to 10.7% at week 42 on the efficacy estimand versus 1.7% on placebo, p&lt;0.001</strong>. <strong>No confidence intervals are reported, no week-28 primary-endpoint figure is stated numerically in either fetched source, and no per-dose results are given against the doses actually administered</strong> (Zealand Pharma; presentation)

Regulatory position: <strong>Not approved anywhere.</strong> Drugs@FDA returns "No matches found!" for petrelintide and the openFDA NDC directory returns <strong>no records at all</strong> (openFDA Drugs@FDA; openFDA NDC). Zero entries in the EMA medicines register generated 03/09/2026 (EMA register). No regulatory filing was located in any fetched source. Anti-doping status: <strong>n.a.</strong>

Safety signal: Discontinuation for an adverse event in <strong>4.9% of placebo (N=81)</strong>, <strong>4.5% of all petrelintide participants (N=404)</strong> and <strong>4.8% of the highest-response dose arm (N=83)</strong>; discontinuation for a gastrointestinal adverse event in <strong>1.5% of placebo and 0.0% of petrelintide</strong> participants; placebo-adjusted nausea <strong>13.4%</strong> (presentation). <strong>Overall treatment-emergent adverse event totals and serious adverse event counts are n.a.</strong> — neither the press release nor the investor presentation states them, and no journal publication was located. <strong>The absence of any peer-reviewed safety report for this compound is itself the finding</strong>

Sources: Zealand Pharma press release ZUPREME-1 conference-call presentation openFDA Drugs@FDA openFDA NDC EMA register

Oral semaglutide 25 mg and 50 mg once daily

Taken as a daily pill, it takes weight off adults with overweight or obesity who do not have diabetes — more than a dummy pill does.

Claim as graded “Reduces body weight, measured on a calibrated scale, relative to placebo in adults with overweight or obesity and without diabetes, taken orally once daily”

Evidence tier B · provenance grade P4 · direction Supported · retail band 1, Strong evidence. Derived by rule 6 - tier A/B with provenance P3+. Group: Hormones and off-label pharmaceuticals. Scope qualifier: body weight

Best available evidence: A <strong>living systematic review and network meta-analysis</strong> of 69 studies and 112,511 participants, searched to October 2025, in which semaglutide is among the interventions reported to have led to the largest weight reductions (living SR/NMA), combined with an <strong>independent regulatory assessment</strong>: FDA approval of oral semaglutide 25 mg under NDA218316 (openFDA Drugs@FDA; Novo Nordisk announcement). Two peer-reviewed phase 3 trials: OASIS 1 (PubMed 37385278; OASIS 1 article PDF) and OASIS 4 (PubMed 40934115; Europe PMC record)

Sample and design: OASIS 1 NCT05035095, phase 3, randomised, double-blind, placebo-controlled, 68 weeks: <strong>667 participants randomised 1:1</strong> to oral semaglutide escalated to 50 mg once daily (334) or placebo (333), both with lifestyle intervention; coprimary endpoints percentage body-weight change and a reduction of at least 5% at week 68 (PubMed 37385278; article PDF). OASIS 4 NCT05564117, 71-week double-blind randomised placebo-controlled trial at 22 sites in four countries: <strong>307 participants randomised 2:1</strong> to oral semaglutide 25 mg once daily (205) or placebo (102), enrolling people without diabetes with a BMI of 30 or higher, or 27 or higher with at least one obesity-related complication; coprimary endpoints at week 64 (Europe PMC record)

Effect as reported: OASIS 1 week 68: <strong>−15.1% (SE 0.5)</strong> with 50 mg versus <strong>−2.4% (SE 0.5)</strong> on placebo, estimated treatment difference <strong>−12.7 percentage points</strong> (95% CI −14.2 to −11.3), <strong>p&lt;0.0001</strong>; reduction of at least 5% in <strong>269 of 317 (85%)</strong> versus <strong>76 of 295 (26%)</strong>, odds ratio 12.6 (95% CI 8.5 to 18.7) (PubMed 37385278; article PDF). OASIS 4 week 64: <strong>−13.6%</strong> with 25 mg versus <strong>−2.2%</strong> on placebo, estimated difference <strong>−11.4 percentage points</strong> (95% CI −13.9 to −9.0), <strong>P&lt;0.001</strong>, with reductions of at least 5%, 10%, 15% and 20% all P&lt;0.001 versus placebo (Europe PMC record). The <strong>lower of the two doses gives the smaller effect and both are published here</strong>; the 25 mg result is the one carrying regulatory assessment

Regulatory position: <strong>The 25 mg dose is approved in the United States; the 50 mg dose is not.</strong> Drugs@FDA lists NDA218316, submission ORIG-1, status "AP" dated 20251222, with tablet strengths of 1.5 mg, 4 mg, 9 mg and 25 mg (openFDA Drugs@FDA). The sponsor states verbatim that "The US Food and Drug Administration (FDA) has approved the Wegovy® pill (once-daily oral semaglutide 25 mg) to reduce excess body weight and maintain weight reduction long term and to reduce the risk of major adverse cardiovascular events", and that a submission was made to the EMA in the second half of 2025 and the product is "currently pending marketing approval from the EMA" (Novo Nordisk). The EMA medicines register generated 03/09/2026 lists Rybelsus and Wegovy among authorised semaglutide products but records <strong>no authorisation for an oral 25 mg or 50 mg weight-management product</strong> (EMA register). Anti-doping status: <strong>n.a.</strong>

Safety signal: OASIS 1, with denominators: adverse events in <strong>307 of 334 (92%)</strong> on oral semaglutide 50 mg versus <strong>285 of 333 (86%)</strong> on placebo; gastrointestinal adverse events in <strong>268 of 334 (80%)</strong> versus <strong>154 of 333 (46%)</strong>, mostly mild to moderate. <strong>Serious adverse event counts and discontinuation-for-adverse-event rates are n.a.</strong> for OASIS 1 — searched the PubMed abstract, the sponsor-hosted article PDF and the Lancet full text, which returned a client error (PubMed 37385278; article PDF). OASIS 4: gastrointestinal adverse events in <strong>74.0% versus 42.2%</strong>, <strong>denominators not stated in the fetched record</strong>, though the arm sizes were 205 and 102 (Europe PMC record)

Sources: living SR/NMA openFDA Drugs@FDA Novo Nordisk announcement PubMed 37385278 OASIS 1 article PDF PubMed 40934115 Europe PMC record EMA register

Tesofensine (NS2330)

Taken as a daily pill for 24 weeks, it takes weight off adults with obesity — more than a dummy pill does.

Claim as graded “Reduces body weight, measured on a calibrated scale, relative to placebo at week 24 in adults with obesity, taken orally once daily”

Evidence tier C · provenance grade P4 · direction Supported · retail band 2, Moderate evidence. Derived by rule 8 - tier C with provenance P3+. Group: Hormones and off-label pharmaceuticals. Scope qualifier: body weight at 24 weeks only; the phase 3 result remains unpublished and no cardiovascular or long-term outcome data were located

Best available evidence: <strong>None for this claim.</strong> No synthesis was located. The only peer-reviewed randomised trial is the <strong>phase 2 TIPO-1 trial</strong> in the Lancet (PubMed 18950853). A <strong>phase 3 registration trial was reported only in a company press release</strong> in 2018 and, per the sponsor, further detail was withheld pending intellectual-property and data-protection rights; <strong>no peer-reviewed publication of it was located</strong> (Saniona phase 3 press release, PDF)

Sample and design: TIPO-1 NCT00394667, phase 2, five centres in Denmark, 24 weeks, BMI 30 to 40: <strong>203 participants randomised</strong> to tesofensine 0.25 mg (52), 0.5 mg (50), 1.0 mg (49) or placebo (52) once daily; <strong>161 of 203 (79%) completed</strong> (PubMed 18950853). Phase 3 "Viking", press release only: <strong>372 patients enrolled and randomised 1:1:1</strong> to oral tesofensine 0.25 mg, 0.50 mg or placebo once daily in a 24-week, randomised, double-blinded, placebo-controlled, three-armed parallel longitudinal trial conducted by Productos Medix in Mexico, with all patients on the same reduced-calorie diet and exercise programme; primary endpoint percentage body-weight change from baseline at week 24 (Saniona press release)

Effect as reported: TIPO-1 week 24, mean weight loss: <strong>4.5% (SE 0.87)</strong> at 0.25 mg, <strong>9.2% (SE 0.91)</strong> at 0.5 mg, <strong>10.6% (SE 0.84)</strong> at 1.0 mg versus <strong>2.0% (SE 0.60)</strong> on placebo, <strong>all p&lt;0.0001</strong>. <strong>No 95% confidence intervals are reported on the fetched page.</strong> Heart rate rose by 7.4 beats per minute at 0.5 mg, p=0.0001 (PubMed 18950853). Phase 3: <strong>"Ten per cent average weight loss in 24 weeks"</strong> with more than half of patients losing more than ten per cent, and p&lt;0.001 in both the intent-to-treat last-observation-carried-forward and completers populations; <strong>no per-arm figures, no confidence intervals and no placebo comparator value are given in the press release</strong> (Saniona press release)

Regulatory position: <strong>Not approved anywhere.</strong> Drugs@FDA returns "No matches found!" for tesofensine; four openFDA NDC records exist and <strong>all four are marketing category and product type "BULK INGREDIENT"</strong>, listed by DARMERICA LLC, Qingdao Biopeptek, Qingdao Qinzeyuan and Cambrex Karlskoga — supplier directory listings, <strong>not an approval</strong> (openFDA Drugs@FDA; openFDA NDC). Zero entries in the EMA medicines register generated 03/09/2026 (EMA register). In Mexico, the sponsor states verbatim that its partner "has not received approval from the Mexican regulatory agency (Cofepris) for tesofensine for the treatment of obesity", and that the partner "is entering a dialogue with the agency regarding the path forward" (Saniona regulatory announcement). Anti-doping status: <strong>n.a.</strong>

Safety signal: <strong>Cardiovascular signal.</strong> In TIPO-1, heart rate rose by <strong>7.4 beats per minute at the 0.5 mg dose, p=0.0001</strong>, and <strong>42 of 203 participants (21%) did not complete the trial</strong>. <strong>Treatment-emergent adverse event rates, serious adverse event counts and discontinuation-for-adverse-event rates with denominators are n.a.</strong> — searched the PubMed abstract, two open-access mirrors of the Lancet paper which returned crawler errors, and the ScienceDirect record (PubMed 18950853). For the phase 3 trial the sponsor states only that adverse-event incidence was low and similar to placebo, with "a low but statically significant increase in heart rate and no significant effect on blood pressure at any of the doses tested", and cites a combined safety database of approximately 1,600 patients across more than 20 trials; <strong>no adverse-event rates, denominators or serious adverse event counts are given</strong>, and this is a press release with no peer review (Saniona press release)

Sources: PubMed 18950853 Saniona phase 3 press release, PDF openFDA Drugs@FDA openFDA NDC EMA register Saniona regulatory announcement

GHRP-6 (growth hormone releasing peptide-6, His-D-Trp-Ala-Trp-D-Phe-Lys-NH2)

Injected into a vein, it raises growth hormone in the blood of healthy younger and older adults.

Claim as graded “Raises circulating growth hormone concentration, measured by serum GH radioimmunoassay, after acute intravenous administration in healthy young and older adults”

Evidence tier D · provenance grade P4 · direction Supported · retail band 6, Too little evidence yet. Derived by rule 4 - tier D or E. Group: SARMs and research peptides. Scope qualifier: acute GH release only; no chronic dosing, body-composition or functional endpoint located

Best available evidence: No synthesis and no randomised placebo-controlled trial located. The best fetched human evidence is a single within-subject pharmacodynamic study in which each participant received three intravenous tests in random order at least one week apart, with no placebo arm (Micic et al., PubMed 7734029)

Sample and design: n=18 total: 9 healthy young adults aged 22 ± 1.1 years and 9 healthy older adults aged 59.5 ± 1.7 years. Three test conditions per subject, order randomised, separated by at least one week, performed at 0900 h after an overnight fast. No placebo condition, no blinding stated, no registration ID (PubMed 7734029)

Effect as reported: 90 µg of GHRP-6 administered as an intravenous bolus. Combined GHRH 100 µg plus GHRP-6 90 µg produced a greater GH rise than GHRH alone in young adults (F=21.9, p<0.001) and in older adults (F=21.8, p<0.001). GH responses to GHRP-6 alone did not differ significantly between young and older subjects (F=0.71, p=NS), whereas responses to GHRH alone were greater in the young (F=3.45, p=0.03). No 95% CIs and no absolute GH concentrations are reported in the fetched source (PubMed 7734029)

Regulatory position: Not an approved medicine in the US or EU in any fetched source. FDA placed GHRP-6 in 503A Category 3, "Bulk Drug Substances Nominated Without Adequate Support" (FDA 503A nominated-substances list, updated 14 May 2026) and in 503B category 2 on September 29, 2023 (FDA category 2 list). WADA 2026 List S2.2.4 Growth hormone releasing factors, naming "GH-releasing peptides (GHRPs) [e.g. alexamorelin, examorelin (hexarelin), GHRP-1, GHRP-2 (pralmorelin), GHRP-3, GHRP-4, GHRP-5 and GHRP-6]" (WADA 2026 List)

Safety signal: FDA: "Compounded drugs containing GHRP-6 may pose risk for immunogenicity for certain routes of administration due to the potential for aggregation and peptide-related impurities. FDA has identified limited safety-related information, but the available data reveal safety concerns including potential effect on cortisol and increase in blood glucose due to decreases in insulin sensitivity" (FDA category 2 list). No adverse-event rates with denominators were located in any fetched source — the absence of published long-term safety data is itself a finding

Sources: Micic et al., PubMed 7734029 FDA 503A nominated-substances list, updated 14 May 2026 FDA category 2 list WADA 2026 List

Sermorelin (GHRH 1-29, GRF 1-29, [Nle27]GHRH-(1-29)-NH2, GEREF)

It raises overnight growth hormone output and a growth-related blood hormone in healthy adults aged 55 to 71.

Claim as graded “Raises 12-hour integrated nocturnal growth hormone secretion and serum IGF-1 in healthy adults aged 55–71, measured by 10-minute-interval venous sampling and serum assay”

Evidence tier C · provenance grade P4 · direction Supported · retail band 2, Moderate evidence. Derived by rule 8 - tier C with provenance P3+. Group: SARMs and research peptides. Scope qualifier: for GH and IGF-1; <strong>Not supported</strong> for bone mineral density and for lean mass in women; IGF-1 elevation was not maintained to 16 weeks

Best available evidence: A single small single-blind randomised placebo-controlled trial, reported across two papers from the same cohort. No meta-analysis and no second independent adult efficacy RCT was located (Khorram, Laughlin & Yen, JCEM 1997, PubMed 9141536; companion immune paper, JCEM 1997;82:3590-6, PubMed 9360512)

Sample and design: n=19 healthy age-advanced adults, 10 women and 9 men, aged 55–71. Single-blind, randomised, placebo-controlled; 4 weeks of nightly saline placebo followed by 16 weeks of active treatment; total duration 5 months; 12-hour nocturnal sampling from 2000–0800 h at 10-minute intervals. No registration ID; the trial predates trial registration (PubMed 9141536)

Effect as reported: 10 µg/kg administered subcutaneously by nightly self-injection at 2100 h for 16 weeks. Twelve-hour integrated nocturnal GH increased versus placebo in women (p<0.01) and men (p<0.05); at 4 weeks integrated GH secretion rose 107% in men and 70% in women (p<0.05), sustained 16 weeks; serum IGF-1 rose 28% in both sexes (p<0.001 at 4 weeks), elevated for 12 weeks and returning toward baseline by 16 weeks. Lean body mass increased in men only (p<0.05); no change in bone mineral density, no change in body weight, no other body-composition change (PubMed 9141536; PubMed 9360512). No 95% CIs and no absolute hormone concentrations reported in either fetched abstract

Regulatory position: <strong>Formerly approved, now withdrawn.</strong> GEREF held NDA 19-863, approved December 28, 1990 (pituitary GH-secretion testing) and NDA 20-443, approved September 26, 1997 (idiopathic GH deficiency in children with growth failure). EMD Serono requested withdrawal of both; withdrawal was announced at 74 FR 23407 and took effect on June 18, 2009. FDA then determined the products "were not withdrawn from sale for reasons of safety or effectiveness" and moved them to the Orange Book "Discontinued Drug Product List" (Federal Register 78 FR 13921, 4 March 2013). Neither approval covered adults. Sermorelin does not appear on the FDA 503A nominated-substances list fetched (FDA 503A list) nor on the category 2 safety-risk list (FDA category 2 list). EU status: `n.a.` — no EMA record was fetched. WADA 2026 List S2.2.4, naming "growth hormone-releasing hormone (GHRH) and its analogues (e.g. CJC-1293, CJC-1295, sermorelin and tesamorelin)" (WADA 2026 List)

Safety signal: The 2013 Federal Register notice states FDA "independently evaluated relevant literature and data for possible postmarketing adverse events for both GEREF products" but reports no event counts, rates or comparative data (Federal Register). Neither fetched trial abstract reports adverse events or discontinuations. No serious harm signal was identified in the fetched sources; the absence of published adult safety data across a 16-week exposure is itself a finding

Sources: Khorram, Laughlin & Yen, *JCEM* 1997, PubMed 9141536 companion immune paper, *JCEM* 1997;82:3590-6, PubMed 9360512 Federal Register 78 FR 13921, 4 March 2013 FDA 503A list FDA category 2 list WADA 2026 List

Semax (ACTH 4-10 analogue, Met-Glu-His-Phe-Pro-Gly-Pro)

It raises scores on standard thinking tests in adults.

Claim as graded “Increases cognitive performance, measured on a validated cognitive test battery, in adults”

Evidence tier E · provenance grade P0 · direction Insufficient · retail band 6, Too little evidence yet. Derived by rule 3 - direction Insufficient. Group: SARMs and research peptides.

Best available evidence: None for this claim. No trial measuring cognitive test performance in humans was located. FDA, evaluating semax on its own initiative in 2026, concluded: "There is insufficient evidence to support the effectiveness for semax (free base) or semax acetate for cerebral ischemia, migraine, or trigeminal neuralgia. Only two available references were available with insufficient details on the design and conduct of the studies, and the available references demonstrated lack of effectiveness" (FDA briefing document, PCAC 23–24 July 2026). An independent academic review states there is "little evidence" of any cognitive benefit from semax in healthy people and "no evidence for Alzheimer's disease" (Cognitive Vitality / ADDF review)

Sample and design: n.a. for cognitive performance. The nearest human data are a single 24-subject brain-imaging study measuring network topography, not cognition: 11 men and 13 women aged 43.9 ± 9.5, 14 receiving semax and 10 placebo, resting-state fMRI before and 5 and 20 minutes after administration; randomisation and blinding are not stated (Lebedeva et al., PubMed 30225715). Total documented human exposure across all fetched FDA-reviewed references: "33-47 healthy adults, 69 adults with medical conditions … and 451 children with either depression or tics/Tourette Syndrome", all intranasal (FDA). No ClinicalTrials.gov interventional record was returned for semax when the registry API was queried in this session (ClinicalTrials.gov API v2, query.intr=semax)

Effect as reported: no numeric effect reported for cognitive performance in humans. The fMRI study reports "a greater volume of the rostral, medial frontal cortex subcomponent of the default mode network" with semax versus controls, with no confidence intervals and no p-values (PubMed 30225715; confirmed as reporting no numerical data by the ADDF review). In the one pain study FDA could evaluate, 4 of 12 migraine subjects reported cessation of pain after 0.5 mg/kg administered intranasally once, and in trigeminal neuralgia the authors concluded "semax does not exhibit analgesic activity by itself" (FDA)

Regulatory position: Not an approved medicine in the US or EU in any fetched source. FDA proposed in July 2026 "that Semax (free base) NOT be included on the 503A Bulks List" and "that Semax acetate NOT be included on the 503A Bulks List" (FDA PCAC overview memo). Semax (heptapeptide) also appears among substances "nominated but withdrawn" on FDA's significant-safety-risk page (FDA). Russian registration status: `n.a.` — no Russian regulatory register page was fetched in this session. Not named in the fetched WADA 2026 List (WADA)

Safety signal: FDA: "We were unable to find literature that discussed administration via subcutaneous injection ROA. The only ROA discussed in the literature was intranasal." One FAERS report (#25343150) was retrieved through 3 December 2025: a consumer reported ocular pain and eye burning after Semax 0.1% nasal drops purchased online, with hospitalisation and pain unresolved a year later. FDA also flags "possible anti-thrombotic properties of semax and this raises concern about the risk of bleeding". Of eight studies tabulated, one reported no adverse events and the rest "AEs were not discussed" (FDA)

Sources: FDA briefing document, PCAC 23–24 July 2026 Cognitive Vitality / ADDF review Lebedeva et al., PubMed 30225715 ClinicalTrials.gov API v2, query.intr=semax FDA PCAC overview memo FDA WADA

Selank (TP-7, tuftsin analogue)

It lowers anxiety scores in adults with generalised anxiety or nervous exhaustion.

Claim as graded “Reduces anxiety, measured on the Hamilton anxiety scale, in adults with generalised anxiety disorder or neurasthenia”

Evidence tier C · provenance grade P2 · direction Insufficient · retail band 6, Too little evidence yet. Derived by rule 3 - direction Insufficient. Group: SARMs and research peptides.

Best available evidence: Two Russian-language randomised trials, both with an active comparator and neither placebo-controlled. No synthesis, no English-language full text, and no numeric effect estimate in any fetched source (Zozulia et al. 2008, PubMed 18454096; Medvedev et al. 2015, PubMed 26356395)

Sample and design: Trial 1: n=62 patients with generalised anxiety disorder and neurasthenia, 30 receiving selank and 32 receiving medazepam; indexed as a Randomized Controlled Trial; blinding not stated, randomisation procedure not described, dose not stated, route not stated, duration not stated (PubMed 18454096). Trial 2: n=70 patients with anxiety-phobic, hypochondriac and somatoform disorders (ICD-10 F40.2-9, F41.1-9, F45.0-2), 30 on phenazepam monotherapy versus 40 on selank plus phenazepam; dose, route and duration not stated (PubMed 26356395). No ClinicalTrials.gov record for selank was returned when the registry API was queried in this session (ClinicalTrials.gov API v2, query.term=selank)

Effect as reported: no numeric effect reported in any fetched source. Trial 1 states only that "the anxiolytic effects of selank and medazepam were similar", with additional antiasthenic and psychostimulant effects; instruments were the Hamilton scale, Zung scale and CGI; no numerical outcome values, no confidence intervals and no p-values are provided (PubMed 18454096). Trial 2 states the phenazepam effect "was achieved earlier when treatment was optimized with selank, according to HDRS", again without numbers, CIs or p-values (PubMed 26356395)

Regulatory position: Not an approved medicine in the US or EU in any fetched source. Selank appears among substances "nominated but withdrawn" as "Selank acetate (TP-7)" on FDA's significant-safety-risk page (FDA) and does not appear on the 503A nominated-substances list fetched (FDA 503A list). Russian registration status: `n.a.` — asserted widely online but no primary register entry was fetched. Not named in the fetched WADA 2026 List (WADA)

Safety signal: FDA: "Compounded drugs containing selank acetate may pose risk for immunogenicity for certain routes of administration due to the potential for aggregation and peptide-related impurities. FDA lacks important information regarding any safety issues raised by selank acetate administered to humans" (FDA). Trial 2 reports that combined treatment decreased phenazepam-related attention impairment, memory impairment, asthenia and sedation, with tolerability rated on the UKU scale, but gives no rates and no denominators (PubMed 26356395). No serious harm signal identified in the fetched sources

Sources: Zozulia et al. 2008, PubMed 18454096 Medvedev et al. 2015, PubMed 26356395 ClinicalTrials.gov API v2, query.term=selank FDA FDA 503A list WADA

Delta sleep-inducing peptide (DSIP, emideltide)

It changes the shape and efficiency of sleep, measured overnight in a sleep lab, in adults with long-term insomnia.

Claim as graded “Alters sleep architecture and sleep efficiency, measured by polysomnography, in adults with chronic insomnia”

Evidence tier C · provenance grade P4 · direction Contested · retail band 4, Experts disagree. Derived by rule 2 - direction Contested. Group: SARMs and research peptides.

Best available evidence: A formal FDA regulatory assessment of the entire published human literature, concluding: "There is insufficient evidence to make a conclusion on the effectiveness of IV emideltide for chronic insomnia. Although some studies showed that short-term IV emideltide appeared to exert an effect on disturbed sleep, the results appear inconclusive and at best preliminary. Clinical studies authored by Schneider-Helmert reported [favourable] sleep outcomes following short-term IV emideltide in subjects with chronic insomnia, whereas in the two DB PC studies by Bes et al. 1992 and Monti et al. 1987 who conducted similarly designed investigations did not reach the same conclusions" (one word replaced in brackets under the lexical rule in §6 of the specification; the original regulator wording is on the linked page) (FDA emideltide briefing document, PCAC 23–24 July 2026)

Sample and design: Every human study identified by FDA was small and intravenous. Largest: n=18 chronic insomniacs (nine aged 29–59, nine aged 60–83), double-blind, 30 nmol/kg administered intravenously each evening for five nights with two placebo baseline nights; and n=14 middle-aged chronic insomniacs, 30 nmol/kg administered intravenously on seven successive nights, which FDA re-classified as an externally matched controlled study because "there was no placebo arm for the insomnia subjects". Others: n=6 randomised double-blind intra-subject placebo-controlled crossover (25 nmol/kg intravenously over 4 minutes before bedtime), n=6 first-in-human crossover, n=6 and n=4 crossover series, n=2 neurotic insomniacs, and n=1 narcolepsy case report (FDA). FDA found no study at all for the nominated subcutaneous route: "There was no study evaluating effectiveness to support use of emideltide (free base) or emideltide acetate via the nominated SC route of administration (ROA) for chronic insomnia." A ClinicalTrials.gov interventional query for delta sleep-inducing peptide executed in this session returned a single record, an unrelated L-carnitine and insulin-resistance study (NCT05251207), and no trial of the peptide itself (ClinicalTrials.gov API v2)

Effect as reported: In the n=14 study, mean sleep-onset latency fell from 58.4 min at baseline to 38.9 min after the first injection and 27.6 min after the final injection; wake after sleep onset from 120.8 min to 78 min to 65.6 min; total sleep time from 313 min to 363.1 min to 385 min; sleep efficiency +27% of baseline. In the n=18 study, elderly baseline sleep latency of 1–2 hours fell to 22–30 minutes, and total sleep time rose +97 min (elderly) versus +57 min (middle-aged) from baselines of 241.6 and 329.1 min. In the n=6 crossover, emideltide "did not influence sleep-onset latency (SOL) or final waking times compared with placebo", with only "tendencies toward a lower NOA and a higher total SE". No 95% CIs and no p-values are reported for any human sleep outcome in the fetched FDA document, and FDA states the comparisons were predominantly within-subject against baseline rather than against a randomised placebo arm (FDA)

Regulatory position: Not an approved medicine in the US or EU in any fetched source. FDA proposed in July 2026 "that Emideltide (free base) NOT be included on the 503A Bulks List" and "that Emideltide acetate NOT be included on the 503A Bulks List" (FDA PCAC overview memo); the substance also appears as "Emideltide (DSIP)" among nominated-but-withdrawn substances (FDA). Not named in the fetched WADA 2026 List (WADA)

Safety signal: FDA: "Compounded drugs containing emideltide may pose risk for immunogenicity for certain routes of administration and may have complexities with regard to peptide-related impurities and API characterization. FDA has not identified safety-related information regarding emideltide for the proposed route of administration. Therefore, the agency lacks sufficient information to know whether the drug would cause harm if administered to humans" (FDA). The only tolerability observation located is that five of six subjects reported "a feeling of pressure to sleep" immediately after infusion (FDA briefing document). No adverse-event rates with denominators were located; every human study is at least 38 years old

Sources: FDA emideltide briefing document, PCAC 23–24 July 2026 ClinicalTrials.gov API v2 FDA PCAC overview memo FDA WADA

GHK-Cu (copper tripeptide-1, glycyl-L-histidyl-L-lysine copper)

Rubbed on the face twice a day for eight weeks, it shrinks wrinkles in women aged 40 to 65, measured by a 3-D scan of the skin.

Claim as graded “Reduces facial wrinkle volume and wrinkle depth, measured by PRIMOS three-dimensional optical profilometry, in women aged 40–65 after 8 weeks of twice-daily topical application”

Evidence tier C · provenance grade P4 · direction Supported · retail band 2, Moderate evidence. Derived by rule 8 - tier C with provenance P3+. Group: SARMs and research peptides. Scope qualifier: wrinkle volume versus vehicle; <strong>Not supported</strong> for wrinkle depth versus the active cosmetic comparator

Best available evidence: One randomised, double-blind, split-face clinical trial with an instrument-measured primary outcome, conducted at a contract cosmetic research institute; plus a set of 12-week facial and periorbital studies that are described only second-hand in a review, without numbers (split-face trial, J Aging Sci; Pickart & Margolina review, Int J Mol Sci)

Sample and design: n=40 female volunteers aged 40–65; 39 completed. Randomised into two groups (Group 1 n=19, Group 2 n=20); split-face design with the GHK-Cu nano-carrier serum applied to the right side of every face and the comparator to the left; identically packaged coded containers; investigators and subjects blinded; 10-day cosmetic-free run-in; twice daily for 8 weeks; comparators were a commercial Matrixyl 3000 product and a serum vehicle without GHK-Cu. The GHK-Cu concentration in the serum is not stated in the publication. Analysis by linear mixed model with random intercept per patient. No trial registration ID (split-face trial)

Effect as reported: Versus vehicle control at 8 weeks: wrinkle volume −0.4 units, relative difference 55.8%, p<0.0001; wrinkle depth −28.3 µm, relative difference 32.8%, p=0.0123. Versus the commercial Matrixyl 3000 comparator: wrinkle volume 31.6%, p=0.0044; wrinkle depth 23.4%, p=0.0577 — not significant at the 0.05 threshold the authors set. Within-arm changes from baseline at 8 weeks were −26.8 ± 12.8% (depth) and −25.8 ± 9.4% (volume) for GHK-Cu versus −15.3 ± 7.2% and −15.0 ± 5.2% for vehicle. No 95% CIs are reported. (split-face trial). The 71-woman 12-week facial-cream study and the 41-woman periorbital study are reported in the review without any numeric result, CI or p-value (PMC6073405)

Regulatory position: <strong>Cosmetic ingredient, not an approved medicine.</strong> FDA removed "GHK-Cu (except for injectable routes of administration)" from 503A category 1 on April 22, 2026 because the nominations were withdrawn; on May 5, 2026 one nominator clarified it was withdrawing only the injectable nomination, and FDA stated "GHK-Cu (except for injectable routes of administration) will be added back to category 1" (FDA 503A list, updated 14 May 2026). The injectable route sits on the significant-safety-risk page (FDA). FDA "intends to consult the Pharmacy Compounding Advisory Committee (PCAC) before the end of February 2027 regarding the potential inclusion of GHK-Cu on the 503A bulks list" (FDA 503A list); GHK-Cu is on the announced agenda for that meeting (FDA PCAC meeting notice). EU status: `n.a.` — no EMA or EU cosmetics-regulation page was fetched. Not named in the fetched WADA 2026 List (WADA)

Safety signal: Topical tolerability in the only randomised trial fetched: 1 of 40 participants experienced minor unwanted skin reactions on both sides of the face, stopped application, and symptoms resolved without medical treatment; 39 of 40 tolerated the 8-week application period (split-face trial). For the injectable route FDA states: "Compounded injectable drugs containing GHK-Cu may pose risk for immunogenicity due to the potential for aggregation and peptide-related impurities. There are limited data in humans to inform safety-related considerations" (FDA)

Sources: split-face trial, *J Aging Sci* Pickart & Margolina review, *Int J Mol Sci* FDA 503A list, updated 14 May 2026 FDA FDA PCAC meeting notice WADA

Thymosin alpha-1 (thymalfasin, Ta1, ZADAXIN)

It cuts the chance of dying within 28 days in adults with sepsis.

Claim as graded “Reduces 28-day all-cause mortality in adults with sepsis”

Evidence tier C · provenance grade P4 · direction Contested · retail band 4, Experts disagree. Derived by rule 2 - direction Contested. Group: SARMs and research peptides.

Best available evidence: Two directly conflicting sources at the top of the evidence hierarchy: a 1,089-patient multicentre double-blind placebo-controlled phase 3 trial that found no difference (TESTS, BMJ 2025), and a 2025 systematic review and meta-analysis of 11 RCTs that found a pooled benefit but detected publication bias and failed its trial-sequential information size (Gu et al., Front Cell Infect Microbiol 2025;15:1673959, PMC12440967)

Sample and design: TESTS: n=1,106 randomised, 1,089 in the modified intention-to-treat analysis (thymosin α1 542, placebo 547), 22 centres in nine Chinese provinces, September 2016 to December 2020, 1:1 computer-generated block randomisation stratified by centre and age, double-blinded to investigators, participants, care providers and statisticians, matched lyophilised-saline placebo, seven days of treatment, 28-day primary endpoint, independent DSMB (BMJ). Meta-analysis: 11 RCTs, 1,927 patients (967 versus 960), all conducted in China, PROSPERO CRD42024628937, searches to 17 January 2025 including CNKI, VIP and Wanfang (PMC12440967)

Effect as reported: Conservative estimate published first. TESTS, 1.6 mg administered subcutaneously every 12 hours for seven days: 28-day mortality 127/542 (23.4%) versus 132/547 (24.1%), HR 0.99, 95% CI 0.77 to 1.27, p=0.93; 90-day mortality 31.0% versus 32.4%, HR 0.94, 95% CI 0.76 to 1.16, p=0.54 (BMJ). In the meta-analysis, the five higher-quality RCTs (n=1,568) gave OR 0.82, 95% CI 0.65 to 1.03, p=0.09, the multicentre subgroup OR 0.86, 95% CI 0.68 to 1.08, p=0.20, and pooled patient-level data from the two high-quality multicentre RCTs HR 0.88, 95% CI 0.72 to 1.08. The headline figure alongside is OR 0.73, 95% CI 0.59 to 0.90, p=0.003 across all 11 RCTs — driven by six lower-quality studies totalling 359 patients (OR 0.41, 95% CI 0.24 to 0.68, p=0.0007; between-quality-subgroup difference p=0.02) and by single-centre studies (OR 0.40, 95% CI 0.25 to 0.63; design-subgroup difference p=0.004). Publication bias was evident on the funnel plot and confirmed by Egger's test, p=0.01, and trial sequential analysis "did not achieve the required information size". No overall I² value is reported (PMC12440967)

Regulatory position: <strong>Not approved in the United States.</strong> FDA: "Ta1 is not approved in the United States, Japan, or Europe (except Italy)"; "FDA has approved no drug products containing Ta1"; "Ta1 is not recognized in the European or Japanese Pharmacopoeias". A website advertising a compounded subcutaneous product as "FDA-approved" is addressed directly: "however, FDA has approved no drug products containing Ta1". On the compounding question FDA concluded: "On balance, the physicochemical characterization, information on historical use, lack of evidence of effectiveness, and safety information identified for both Ta1 (free base) and Ta1 acetate weigh against them being added to the 503A Bulks List" (FDA briefing document, PCAC 4 December 2024). Ta1 appears among nominated-but-withdrawn substances (FDA). Not named in the fetched WADA 2026 List (WADA)

Safety signal: In the phase 3 trial, adverse events did not differ: at least one adverse event 360/542 (66.4%) versus 370/547 (67.6%), difference −1.2%, 95% CI −6.8% to 4.4%, p=0.70; serious adverse events 145/542 (26.8%) versus 160/547 (29.3%), difference −2.5%, 95% CI −7.8% to 2.8%, p=0.38; commonest events anaemia 10.7%, fever 9.6%, abdominal distension 5.4%, coagulation disorders 4.8%; "no unexpected serious adverse events related to thymosin α1 occurred" (BMJ). A pre-specified-threshold caution applies to the subgroup signal of harm in those aged under 60 (HR 1.67, 95% CI 1.04 to 2.67, interaction p=0.01) against a Bonferroni threshold of p=0.006 (BMJ). FDA separately notes "insufficient safety information on use of the substances and a lack of information about immunogenicity risks", no identified carcinogenicity studies, and non-US labels carrying warnings in children, pregnancy and lactation, autoimmune disease and immunosuppressed populations (FDA)

Sources: TESTS, *BMJ* 2025 Gu et al., *Front Cell Infect Microbiol* 2025;15:1673959, PMC12440967 FDA briefing document, PCAC 4 December 2024 FDA WADA

MOTS-c (mitochondrial open reading frame of the 12S rRNA-c, mitochondrial-derived peptide)

Given as an injection, it makes the body handle insulin better or lets people exercise harder.

Claim as graded “Raises insulin sensitivity or exercise capacity in humans following administration of exogenous MOTS-c”

Evidence tier E · provenance grade P0 · direction Insufficient · retail band 6, Too little evidence yet. Derived by rule 3 - direction Insufficient. Group: SARMs and research peptides.

Best available evidence: None. Zero humans have ever received MOTS-c in any published study. FDA, in its 2026 evaluation: "The nomination did not include, and FDA has not identified, any clinical studies or human exposure data for MOTS-c via any route of administration"; "We performed our own search of published medical literature and did not identify clinical studies evaluating administration of MOTS-c-related BDSs in human subjects"; "There is a lack of evidence to evaluate the effectiveness of the MOTS-c-related BDSs for the nominated uses … these studies were limited to in-vitro and in-vivo rodent models" (FDA MOTS-c briefing document, PCAC 23–24 July 2026)

Sample and design: n.a. — zero human participants have received MOTS-c for any endpoint. The first interventional human trial began enrolling in 2026 and has posted no results: NCT07505745, a trial of MOTS-c in adults with prediabetes and overweight or obesity with insulin sensitivity as its subject (registry title on the linked record), phase 2, two-arm multicentre randomised double-blind placebo-controlled parallel-group, quadruple masking, estimated enrolment 120, subcutaneous, 12 weeks of treatment, primary endpoint change in Matsuda index from a 75 g OGTT, actual start 2 February 2026, estimated primary completion 14 February 2027, status Recruiting, "No results posted" (NCT07505745). Human studies that mention MOTS-c measure endogenous peptide concentrations after exercise; none administers it

Effect as reported: no numeric effect reported — no metabolic or exercise-capacity outcome has ever been measured in a human given MOTS-c (FDA; NCT07505745)

Regulatory position: Not an approved medicine anywhere in any fetched source. FDA proposed in July 2026 "that MOTS-c (free base) NOT be included on the 503A Bulks List" and "that MOTS-c acetate NOT be included on the 503A Bulks List" (FDA PCAC overview memo); "Accordingly, we propose not adding MOTS-c (free base) or MOTS-c acetate to the 503A Bulks List" (FDA MOTS-c briefing). Listed as "MOTs-C" among nominated-but-withdrawn substances (FDA). WADA 2026 List S4.4.1 Metabolic modulators, naming "Activators of the AMP-activated protein kinase (AMPK), e.g. … mitochondrial open reading frame of the 12S rRNA-c (MOTS-c)" (WADA 2026 List)

Safety signal: FDA: "There is a lack of clinical and nonclinical safety information on the use of MOTS-c. FDA is particularly concerned about the lack of human data on drug products containing this substance administered via any route of administration. MOTS-c is a peptide containing 16 amino acids for which there is a lack of information to assess its immunogenic safety risk. Therefore, potential safety risks associated with the use of MOTS-c in humans are unknown." FDA identified no acute toxicity, no repeat-dose toxicity, no genotoxicity, no developmental and reproductive toxicity and no carcinogenicity studies of either salt form (FDA). The absence of any published human safety data is the finding

Sources: FDA MOTS-c briefing document, PCAC 23–24 July 2026 NCT07505745 FDA PCAC overview memo FDA WADA 2026 List

Elamipretide (SS-31, MTP-131, bendavia, marketed as Forzinity)

After 24 weeks it lets adults with an inherited muscle disease walk further in six minutes.

Claim as graded “Increases distance walked on the 6-minute walk test at 24 weeks in adults with genetically confirmed primary mitochondrial myopathy”

Evidence tier C · provenance grade P4 · direction Not supported · retail band 5, Evidence says no. Derived by rule 1 - direction Not supported. Group: SARMs and research peptides. Scope qualifier: primary mitochondrial myopathy, 6-minute walk distance

Best available evidence: A single pivotal phase 3 randomised double-blind placebo-controlled trial, MMPOWER-3, which missed both co-primary endpoints. No pooled synthesis exists (Karaa et al., Neurology 2023;101(3):e238-e252)

Sample and design: n=296 screened, 218 randomised 1:1 (elamipretide 109, placebo 109) at 27 centres in Canada, Denmark, England, Germany, Hungary, Italy and the United States; enrolled October 2017 to December 2019; 24-week parallel-group double-blind period with identical glass vials, stratified by sequence alteration; pharmacists, investigators, sponsor and participants blinded; 205 (94%) completed; per-protocol population 198/218 (90.8%). Registered as NCT03323749 (Neurology)

Effect as reported: 40 mg administered subcutaneously once daily for 24 weeks. Co-primary endpoints, intention-to-treat: 6-minute walk distance −3.2 m, 95% CI −18.7 to 12.3, p=0.69; PMMSA total fatigue score −0.07, 95% CI −0.69 to 0.54, p=0.81. Per-protocol: −2.2 m, 95% CI −16.9 to 12.5, p=0.77 and 0.09, 95% CI −0.54 to 0.72, p=0.78. In the pre-specified mtDNA subgroup the 6MWT point estimate favoured placebo (−11.0 m, 95% CI −28.1 to 6.1, p=0.21); in the smaller nDNA subgroup (n=29 versus n=29) it favoured elamipretide (25.2 m, 95% CI 3.1 to 47.3, p=0.03), an unreplicated subgroup finding. The trial did not meet its primary endpoints. No I² is reported; this is a single trial (Neurology)

Regulatory position: <strong>Rejected, then approved under accelerated approval for a different disease.</strong> In Barth syndrome, FDA's advisory-committee briefing stated "The primary endpoint results of SPIBA-201, Part 1, an AWC clinical investigation, do not show a treatment effect of elamipretide versus placebo" and "These echocardiographic data do not show a treatment effect of elamipretide on cardiac structure or function after 12 weeks of treatment", adding "FDA did not agree with the design of SPIBA-001 to evaluate treatment effect of elamipretide" (FDA briefing document, 10 October 2024). FDA issued a complete response action letter dated May 15, 2025, answered by a sponsor amendment of 15 August 2025, and then approved NDA 215244 (Forzinity, 280 mg/3.5 mL) on September 19, 2025 under "NDA 215244 ACCELERATED APPROVAL" pursuant to section 506(c) and 21 CFR 314.510, with an indication naming muscle strength "in adult and pediatric patients with Barth syndrome weighing at least 30 kg", with a required confirmatory randomised placebo-controlled trial (study initiation 03/2026, final report 03/2030) and the standard condition "If your required postmarketing clinical trial fails to verify clinical benefit … we may withdraw this approval" (FDA approval letter; FDA press announcement). There is no approval for primary mitochondrial myopathy. EU status: `n.a.` — no EMA record was fetched. Not named in the fetched WADA 2026 List (WADA)

Safety signal: Near-universal injection-site reactions. In MMPOWER-3: any adverse event 107/109 (98.2%) versus 83/109 (76.1%); treatment-related adverse events 106 (97.2%) versus 56 (51.4%); serious adverse events 5 (4.6%) versus 3 (2.8%), none treatment related; discontinuation for adverse events 8 (7.3%) versus 2 (1.8%); treatment interruption 13 (11.9%) versus 5 (4.6%); no deaths. Injection-site erythema 86.2% versus 28.4%, pruritus 75.2% versus 9.2%, pain 39.4% versus 18.3%, swelling 38.5% versus 6.4%, induration 28.4% versus 5.5%, urticaria 12.8% versus 0 (Neurology). FDA required post-marketing carcinogenicity work, stating "an analysis of spontaneous postmarketing adverse events … will not be sufficient to identify an unexpected serious risk of carcinogenicity and drug-drug interactions", mandating a 26-week TgRasH2 mouse study and a 2-year rat study (FDA approval letter)

Sources: Karaa et al., *Neurology* 2023;101(3):e238-e252 FDA briefing document, 10 October 2024 FDA approval letter FDA press announcement WADA

5-amino-1MQ (5-amino-1-methylquinolinium, NNMT inhibitor)

It causes fat loss, or some other measurable change in how the body handles energy, in people.

Claim as graded “Produces fat loss or a measurable metabolic benefit in humans”

Evidence tier E · provenance grade P0 · direction Insufficient · retail band 6, Too little evidence yet. Derived by rule 3 - direction Insufficient. Group: SARMs and research peptides.

Best available evidence: None. No human being has been given this compound in any published or registered study. A PubMed search executed in this session for the term "5-amino-1MQ" returned zero records (PubMed E-utilities esearch, term=5-amino-1MQ); a search for the full chemical name "5-amino-1-methylquinolinium" returned three records, all preclinical (PubMed esearch): a bladder-cancer study using "5-Amino-1-methylquinolinium iodide" in mouse models (PubMed 39067875) and a diet-induced-obese mouse microbiome study (PubMed 35013352)

Sample and design: n.a. — zero human participants. ClinicalTrials.gov API v2 queries executed in this session for "5-amino-1MQ" and for "5-amino-1-methylquinolinium" each returned an empty study list (ClinicalTrials.gov API v2, query.term=5-amino-1MQ; query.term=5-amino-1-methylquinolinium). The compound also does not appear anywhere on FDA's 503A nominated-substances list (FDA 503A list) or on the category 2 significant-safety-risk list (FDA category 2 list)

Effect as reported: no numeric effect reported in humans. The only weight-related result located is in mice: NNMT inhibition combined with a low-fat diet "promoted dramatic whole-body adiposity and weight loss" in diet-induced obese mice and normalised these measures to age-matched lean animals, with no sample size, no dose, no route, no duration, no confidence intervals and no p-values stated (PubMed 35013352)

Regulatory position: Not an approved medicine anywhere in any fetched source, and not a peptide. It is not under FDA compounding evaluation at all: it appears on neither the 503A nominated-substances list nor the category 2 list (FDA 503A list; FDA category 2 list). Anti-doping: not named in the fetched WADA 2026 List (WADA 2026 List)

Safety signal: No human safety data exist. No adverse-event report, no case report, no tolerability study and no pharmacokinetic study in humans was located in any fetched source; PubMed returns three preclinical records in total for the compound (PubMed esearch). The complete absence of any human safety measurement is the primary finding

Sources: PubMed E-utilities esearch, term=5-amino-1MQ PubMed esearch PubMed 39067875 PubMed 35013352 ClinicalTrials.gov API v2, query.term=5-amino-1MQ query.term=5-amino-1-methylquinolinium FDA 503A list FDA category 2 list WADA 2026 List

Andarine [S-4, GTx-007]

It adds muscle and strips fat off healthy adults who lift weights.

Claim as graded “Increases lean muscle mass and reduces fat mass in healthy adults doing resistance training”

Evidence tier E · provenance grade P0 · direction Insufficient · retail band 6, Too little evidence yet. Derived by rule 3 - direction Insufficient. Group: Prohormones, precursors and peptides.

Best available evidence: No human efficacy study located for this claim. Nearest evidence is a castrated-rat study: S-4 at 3 or 10 mg/kg for 8 wk "restored soleus muscle mass and strength and levator ani muscle mass to that seen in intact animals" (Gao et al., Endocrinology, PMC2039881)

Sample and design: n.a. — no human trial found. ClinicalTrials.gov API v2 returned 0 registered studies with andarine as intervention (ClinicalTrials.gov API). Rat study used male Sprague Dawley rats, n = 7–8 per group, 8 wk daily subcutaneous dosing (PMC2039881)

Effect as reported: No numeric effect reported in humans. Rat data only, reported as mean ± SD with ANOVA at P < 0.05; "Confidence intervals were not reported" (PMC2039881)

Regulatory position: Unapproved new drug. FDA warning letter to TITAN SARMS LLC (MARCS-CMS 719645, December 12, 2025) names "'S-4' (also referred to on your website as Andarine)" and states these products "are unapproved new drugs under section 505(a)" and that marketing them "violates sections 301(d) and 505(a) of the FD&C Act" (FDA warning letter). LiverTox lists "Dose Range used in Clinical Trials: Not reported" for andarine (LiverTox NBK619971)

Safety signal: FDA: "Life-threatening reactions, including liver toxicity, have occurred in people taking products containing SARMs. SARMs also have the potential to increase the risk of heart attack and stroke" (FDA warning letter). FDA consumer alert lists psychosis/hallucinations, sexual dysfunction, "Liver injury and acute liver failure", infertility, pregnancy miscarriage and testicular shrinkage (FDA safety communication). Class LiverTox likelihood score B, "Likely cause of clinically apparent liver injury with jaundice"; initial total bilirubin "generally 4.0 to 8.0 mg/dL", aminotransferases "2 to 5 times the upper limit of normal", severe cases with bilirubin "above 30 mg/dL" can develop renal dysfunction requiring temporary dialysis (LiverTox NBK619971). No andarine-specific case report was found in the 15-manuscript / 18-case systematic review (Eur J Clin Pharmacol, PMC10847181; PMC10204391)

Sources: PMC2039881 FDA warning letter 719645 FDA SARMs safety communication LiverTox NBK619971 PMC10204391 CT.gov API

YK-11 [myostatin inhibitor]

It blocks the protein that limits muscle growth, so it builds more muscle and strength than other SARMs do in weight-trained adults.

Claim as graded “Inhibits myostatin and therefore produces greater muscle hypertrophy and faster strength gains than other SARMs in resistance-trained adults”

Evidence tier E · provenance grade P0 · direction Insufficient · retail band 6, Too little evidence yet. Derived by rule 3 - direction Insufficient. Group: Prohormones, precursors and peptides.

Best available evidence: No human study of any kind located. LiverTox: "Other less well described SARMs that have been developed and have not been approved for use in humans include GSK-2881078, S-23, YK-11, and LGD-3303" (LiverTox NBK619971)

Sample and design: n.a. — ClinicalTrials.gov API v2 returned 0 registered studies with YK-11 as intervention (CT.gov API)

Effect as reported: No numeric effect reported. The hypertrophy claim is quoted by FDA directly from vendor labelling — "YK-11's ability to down regulate myostatin expression may allow for greater muscle hypertrophy" — and FDA treats it as evidence of unapproved drug intent, not as substantiated effect (FDA warning letter)

Regulatory position: Unapproved new drug. FDA warning letter to TITAN SARMS LLC (December 12, 2025) names "'YK-11' (also referred to on your website as Myostatin Inhibitor)" and finds it is "not generally recognized as safe and effective (GRASE)" and is an unapproved new drug (FDA warning letter). Never approved for human use (LiverTox NBK619971)

Safety signal: Stroke report with FDA laboratory confirmation. FDA: "An FDA laboratory recently tested GE Labs Ykarine. The testing found the product contained undeclared trendione… The laboratory analysis also confirmed the product contains YK-11, which is a listed ingredient on the product label. The agency testing was based on an adverse event report of a consumer who used the product and subsequently suffered a stroke. The agency urges consumers not to use GE Labs Ykarine" (FDA safety communication). YK11 appears in one published drug-induced-liver-injury case (Lee et al., alongside LGD-4033 and RAD-140) within a review of 18 cases whose mean peak total bilirubin was 28.5 ± 13.4 mg/dL, mean peak ALT 226.3 ± 142 IU/L and mean peak alkaline phosphatase 283.1 ± 160.9 IU/L (PMC10204391). Class hepatotoxicity as for P-01 (LiverTox NBK619971)

Sources: FDA safety communication FDA warning letter 719645 LiverTox NBK619971 PMC10204391 CT.gov API

S-23

It adds muscle, thickens bone and strips fat off adults.

Claim as graded “Increases lean mass and bone mineral density and reduces fat mass in adults”

Evidence tier E · provenance grade P0 · direction Insufficient · retail band 6, Too little evidence yet. Derived by rule 3 - direction Insufficient. Group: Prohormones, precursors and peptides.

Best available evidence: No human efficacy study located. Nearest evidence is a rat male-contraception study: "S-23 increased bone mineral density and lean mass but reduced fat mass in a dose-dependent manner" (Jones et al., Endocrinology, PMC2630904). Human exposure data exist only as doping-control excretion pharmacology with no efficacy endpoint (Drug Test Anal, PMC12023825)

Sample and design: n.a. for the claim. Rat study: male Sprague Dawley rats, castrated-rat arm n = 5/group, intact-rat arm 42 rats in 7 groups of 6, daily subcutaneous dosing for 14 or 70 d (PMC2630904). Human excretion study: "The fortified drinking yoghurt was administered orally to five healthy male volunteers each" at single doses of 1, 10 and 50 μg and at 5 consecutive daily doses of 1, 10 and 50 μg; urine only, 30 d collection; ethics approval German Sport University Cologne 173/2023 (PMC12023825). ClinicalTrials.gov returned 0 S-23 SARM studies (the 4 hits for the string "S-23" are unrelated anaesthesia, transplant, dental and xenon trials) (CT.gov API)

Effect as reported: No numeric effect reported in humans. Rat binding affinity "inhibitory constant = 1.7 ± 0.2 nm"; ED50 in prostate and levator ani "0.43 and 0.079 mg/d"; "four of six animals showed no sperm in the testis and zero pregnancies (none of six)"; fertility fully reversible with "a 100% pregnancy rate observed after 100 d of recovery" (PMC2630904). Human study reported only urinary detection windows, e.g. intact S-23 detected for 544 h after a single 50 μg dose with average peak 1790 pg/mL at 18.5 h (PMC12023825)

Regulatory position: Not approved anywhere that could be sourced; LiverTox lists S-23 among SARMs that "have not been approved for use in humans" (LiverTox NBK619971). FDA's SARM enforcement posture and class alert apply to products marketed as SARMs generally (FDA safety communication)

Safety signal: No S-23-specific human adverse-event report was found. The published human exposures were microgram-level doping-control administrations that reported no safety outcomes (PMC12023825). Rat data document suppression of spermatogenesis to zero sperm in 4 of 6 animals and suppression of LH by more than 50% (PMC2630904). Class hepatotoxicity signal: LiverTox likelihood score B, bilirubin above 30 mg/dL in severe cases with dialysis-requiring renal dysfunction (LiverTox NBK619971); FDA lists acute liver failure, heart attack, stroke, psychosis, infertility and miscarriage for SARM products (FDA safety communication)

Sources: PMC2630904 PMC12023825 LiverTox NBK619971 FDA safety communication CT.gov API

Androstenedione ["andro", 4-androstene-3,17-dione]

It raises testosterone in the blood and so adds muscle size and strength in young men who lift weights.

Claim as graded “Raises serum testosterone and thereby enhances muscle size and strength gains in young men undertaking resistance training”

Evidence tier C · provenance grade P3 · direction Not supported · retail band 5, Evidence says no. Derived by rule 1 - direction Not supported. Group: Prohormones, precursors and peptides.

Best available evidence: King et al., "Effect of oral androstenedione on serum testosterone and adaptations to resistance training in young men: a randomized controlled trial", JAMA 1999;281(21):2020-8 (PubMed 10359391); corroborated by Broeder et al., Arch Intern Med 2000 (PubMed 11074738)

Sample and design: n = 30 healthy, normotestosterogenic men aged 19–29; "Eight-week randomized controlled trial"; 20 subjects performed 8 wk of whole-body resistance training and were randomised to androstenedione 300 mg/d (n = 10) or placebo (n = 10) (PubMed 10359391). Second trial: n = 50 men aged 35–65, double-blind randomised to placebo, androstenediol 200 mg/d or androstenedione 200 mg/d during 12 wk of high-intensity resistance training (PubMed 11074738)

Effect as reported: "Serum free and total testosterone concentrations were not affected by short- or long-term androstenedione administration." Knee extension strength increased "significantly (P<.05) and similarly in the placebo (770 [55] N vs 1095 [52] N) and androstenedione (717 [46] N vs 1024 [57] N) groups"; type 2 fibre cross-sectional area androstenedione 4703 [471] → 5307 [604] mm2 vs placebo 5271 [485] → 5728 [451] mm2. Estradiol rose to 310 [20] pmol/L at 2 wk, 300 [30] at 5 wk and 280 [20] at 8 wk vs baseline 220 [20] pmol/L, P<.05. Values are mean [SEM]; no 95% confidence intervals were reported (PubMed 10359391). Broeder: total testosterone "significantly increased 16% after 1 month of use, but by the end of 12 weeks… returned to pretreatment levels"; "Neither androstenediol nor androstenedione enhanced the adaptations to resistance training compared with placebo for body composition or muscular strength" (PubMed 11074738)

Regulatory position: US Schedule III controlled substance. The Anabolic Steroid Control Act of 2004 (Public Law 108-358) lists at clause (iv) "androstenedione— (I) 1-androstenedione ([5α]-androst-1-en-3,17-dione); (II) 4-androstenedione (androst-4-en-3,17-dione); and (III) 5-androstenedione (androst-5-en-3,17-dione)" (PLAW-108publ358). FDA: "4-androstenedione is an anabolic steroid, classified by the Drug Enforcement Administration (DEA) as a Schedule III controlled substance" and its use in food "does not meet the criteria for general recognition of safety" (FDA scientific memorandum). DEA: the 1990 and 2004 Acts "placed a total of 64 anabolic steroids in schedule III"; an August 1, 2023 final rule moved the listed steroids to 21 CFR 1308.13(f) (DEA Diversion Control). HHS announced an androstenedione crackdown on March 11, 2004 (FDA warning-letter index)

Safety signal: Adverse lipid shift in both RCTs. King: HDL cholesterol fell from 1.09 [0.08] mmol/L (42 [3] mg/dL) to 0.96 [0.08] mmol/L (37 [3] mg/dL) after 2 wk, P<.05, "and remained low after 5 and 8 weeks" (PubMed 10359391). Broeder: "both androstenediol and androstenedione supplementation adversely affected high-density lipoprotein cholesterol (HDL-C) levels, coronary heart disease risk (representing a 6.5% increase)" with lipid-ratio change diol +5.2%, dione +10.5%, P = .05, while placebo HDL-C rose 5.1% (PubMed 11074738). FDA: "the available reports underscore its potential for serious reproductive, developmental, and cardiovascular toxicity, as well as carcinogenic potential" (FDA memorandum)

Sources: PubMed 10359391 PubMed 11074738 PLAW-108publ358 FDA 4-androstenedione memorandum DEA anabolic steroids factsheet FDA warning-letter index

Epiandrosterone [3beta-androsterone, 3β-hydroxy-5α-androstan-17-one]

Taken by mouth, it adds strength, hardness and sex drive in adults, by converting into a stronger androgen.

Claim as graded “Oral prohormone that increases strength, muscle hardness and libido in adults, acting as a dihydrotestosterone precursor”

Evidence tier E · provenance grade P0 · direction Insufficient · retail band 6, Too little evidence yet. Derived by rule 3 - direction Insufficient. Group: Prohormones, precursors and peptides.

Best available evidence: No human supplementation study located. The NCATS Inxight record reproduces the marketing claim — "Epiandrosterone is used as an anabolic agent (dietary supplement, a precursor to dihydrotestosterone) to increase strength, muscle hardness and also improves libido" — while its curator comment states "No human data available" and the record lists Approval Year: Unknown, Conditions: Unknown with no approved indication (NCATS Inxight 8TR252Z538)

Sample and design: n.a. — no human efficacy trial found. ClinicalTrials.gov returned 3 records matching "epiandrosterone", all dehydroepiandrosterone (DHEA) or adrenal-androgen assay studies (NCT02151006 DHEA in IVF, NCT05811507, NCT05903716), none an epiandrosterone supplementation trial (CT.gov API)

Effect as reported: No numeric effect reported for the claim. The one human study surfaced under this heading tested a different product (astaxanthin plus Saw Palmetto berry lipid extract, "Alphastat® (Mytosterone™)", 42 men aged 37–70) and did not administer epiandrosterone at all (J Int Soc Sports Nutr, PubMed 18700016)

Regulatory position: No approval or regulatory decision could be sourced; the NCATS record shows no approved indication and no approval authority (NCATS Inxight). Epiandrosterone (3β-hydroxy-5α-androstan-17-one) does not appear in the list of anabolic steroids added by the Anabolic Steroid Control Act of 2004 (PLAW-108publ358) and did not appear in the enumerated clause list of the Designer Anabolic Steroid Control Act of 2014, which instead added a catch-all subparagraph for substances "substantially similar" to listed anabolic steroids marketed to promote muscle growth (PLAW-113publ260). Whether that catch-all captures epiandrosterone in a given product is a determination not made in any page fetched here — Not checked

Safety signal: No epiandrosterone-specific human safety data were located: no clinical trial, no case report, no regulator adverse-event finding. The NCATS record contains no human clinical safety data (NCATS Inxight). The only sourced safety inference is by class: FDA concluded of the closely related steroid precursor 4-androstenedione that the available reports "underscore its potential for serious reproductive, developmental, and cardiovascular toxicity, as well as carcinogenic potential" (FDA memorandum) — this is class-level and not epiandrosterone-specific

Sources: NCATS Inxight 8TR252Z538 PubMed 18700016 PLAW-108publ358 PLAW-113publ260 FDA 4-androstenedione memorandum CT.gov API

Melanotan II [MT-II, MII]

It tans the skin without any sun or sunbed exposure.

Claim as graded “Produces skin tanning (increased cutaneous pigmentation) in adults without ultraviolet exposure”

Evidence tier D · provenance grade P2 · direction Supported · retail band 6, Too little evidence yet. Derived by rule 4 - tier D or E. Group: Prohormones, precursors and peptides. Scope qualifier: pilot-scale only; n = 3, no inferential statistics

Best available evidence: Dorr et al., "Evaluation of melanotan-II, a superpotent cyclic melanotropic peptide in a pilot phase-I clinical study", Life Sci 1996;58(20):1777-84 (PubMed 8637402)

Sample and design: "A single-blind, alternating day (saline or MT-II), placebo-controlled trial was conducted in 3 normal male volunteers"; subcutaneous injections Monday–Friday for 2 consecutive weeks; starting dose 0.01 mg/kg, escalated by 0.005 mg/kg increments to 0.03 mg/kg in two subjects and 0.025 mg/kg in one (PubMed 8637402)

Effect as reported: "Two subjects had increased pigmentation in the face, upper body and buttock, as measured by quantitative reflectance and by visual perception 1 week after MT-II dosing ended." Authors' conclusion: "These results demonstrate that MT-II has tanning activity in humans given only 5 low doses every other day by subcutaneous injection." No p-value and no confidence interval were reported; the effect is 2 of 3 subjects in an uncontrolled-for-multiplicity pilot (PubMed 8637402)

Regulatory position: Unapproved new drug in the US. FDA debarment record: "On or about August 30, 2007, Melanocorp received a warning letter from the FDA… The warning letter noted that, based on information and statements on the Melanocorp website, MII constituted a new drug under the FDCA that could not be introduced or delivered for introduction into interstate commerce without an FDA approved application" (FDA media/130124). One registered trial exists and it is not for tanning: NCT07437560, Phase 2, RECRUITING, enrolment 60, "Melanotan II (MT-II) as an Adjunct to NB-UVB Phototherapy for Repigmentation in Stable Nonsegmental Vitiligo", no results posted (CT.gov NCT07437560). A review notes "Multiple national health organizations have issued safety warnings regarding the use of melanotan I and II" (Int J Dermatol review)

Safety signal: Melanoma. A 42-year-old white woman with phototype III skin developed, "progressively over a 3‐month period after subcutaneous injections of melanotan II", enlargement and colour change of an abdominal naevus; histology "revealed a melanoma", Breslow thickness 0·30 mm, Clark level II, Ki-67 "estimated at 10%", with S100a, HMB45, Melan-A, NKIC3 and tyrosinase positive; "New naevi had appeared and other pre‐existing naevi presented enlargement and/or darkening" (Br J Dermatol 2011). Review: "Four case reports have described melanomas emerging from existing moles either during or shortly after the use of melanotan" (Int J Dermatol). Renal infarction. Previously healthy 45-year-old man, right-sided renal infarction "affecting approximately 50% of the kidney" after 27 mg of melanotan II self-administered subcutaneously over 6 months obtained via a web shop; blood pressure 165/95 mmHg in a previously normotensive man, CRP 152 mg/L; hypertension persisted and renal function was 81 mL/min per 1.73 m² at 2 months; "most likely attributed to Melanotan II" (CEN Case Rep, PMC7148395). Rhabdomyolysis. Nelson, Bryant and Aks, "Melanotan II injection resulting in systemic toxicity and rhabdomyolysis", Clin Toxicol (Phila) 2012;50(10):1169–1173 — cited in the renal-infarction paper as prior description of "Melanotan II inducing rhabdomyolysis and renal failure… with sympathomimetic excess causing renal dysfunction" (PMC7148395). Phase I dose-limiting effects: "The 0.03 mg/kg dose produced Grade II somnolence and fatigue in one of two subjects (WHO standards)", mild nausea at most levels, and spontaneous penile erections lasting 1–5 h (PubMed 8637402). No death was reported in any page fetched

Sources: PubMed 8637402 Br J Dermatol 2011 Int J Dermatol review PMC7148395 FDA media/130124 CT.gov NCT07437560

Bremelanotide [PT-141]

It treats erection problems in men.

Claim as graded “Treats erectile dysfunction in men”

Evidence tier C · provenance grade P1 · direction Supported · retail band 3, Widely studied, poorly measured. Derived by rule 5 - provenance cap at P1 or below. Group: Prohormones, precursors and peptides. Scope qualifier: single trial, intranasal formulation, sildenafil non-responders only

Best available evidence: Safarinejad & Hosseini, "Salvage of sildenafil failures with bremelanotide: a randomized, double-blind, placebo controlled study", J Urol 2008;179(3):1066-71 (PubMed 18206919)

Sample and design: "A total of 342 married men (28 to 59 years old) with erectile dysfunction who did not respond to sildenafil were randomly assigned to receive 10 mg bremelanotide as an intranasal spray (group 1, 172)… or a similar regimen of placebo (group 2, 170)"; randomised, double-blind, placebo-controlled; outcomes by self-administered International Index of Erectile Function; patients asked to use at least 16 doses/attempts at home (PubMed 18206919)

Effect as reported: "Positive clinical results were seen in 51 (33.5%) patients in the bremelanotide group compared with 13 (8.5%) patients in the placebo group (p = 0.03)." Greater intercourse satisfaction, p = 0.03. No confidence interval was reported. Authors' own hedge: "Further studies with different dosages and treatment regimens are necessary to draw final conclusions on the efficacy of this drug in erectile dysfunction" (PubMed 18206919)

Regulatory position: Approved — but for a different indication and a different population than the claim. VYLEESI (bremelanotide injection) was approved as NDA 210557 on 06/21/2019 to AMAG Pharmaceuticals, 1.75 mg/0.3 mL subcutaneous (FDA approval package). Approved indication: "VYLEESI is indicated for the treatment of premenopausal women with acquired, generalized hypoactive sexual desire disorder (HSDD)". Label Limitations of Use state expressly: "VYLEESI is not indicated for the treatment of HSDD in postmenopausal women or in men" and "VYLEESI is not indicated to enhance sexual performance" (Vyleesi prescribing information). The erectile-dysfunction claim graded here is therefore an unapproved use of an approved molecule, in a population the label excludes. ClinicalTrials.gov holds no registered bremelanotide/PT-141 trial in erectile dysfunction (CT.gov API)

Safety signal: Contraindicated "in patients who have uncontrolled hypertension or known cardiovascular disease". "VYLEESI transiently increases blood pressure and reduces heart rate after each dose… maximal increases of 6 mmHg in systolic blood pressure (SBP) and 3 mmHg in diastolic blood pressure (DBP)… reduction in heart rate up to 5 beats per minute." Focal hyperpigmentation, including face, gingiva and breasts, in 1% at up to 8 doses/month, but "38% of patients developed focal hyperpigmentation after receiving VYLEESI daily for 8 days; among patients who continued VYLEESI for 8 more consecutive days, an additional 14% developed new focal pigmentary changes… Resolution of the focal hyperpigmentation was not confirmed in all patients after discontinuation." Nausea 40.0% vs 1.3% placebo, flushing 20.3% vs 0.3%, injection site reactions 13.2% vs 8.4%, headache 11.3% vs 1.9%; discontinuation for adverse reactions 18% vs 2% (Vyleesi PI). Development-programme review across 3500 subjects in 43 completed studies: "There were no deaths; a few subjects experienced serious AEs" (J Womens Health 2022, PubMed 35147466). Non-reporting: NCT04943068, Phase 3 bridging trial in premenopausal women with HSDD, enrolment 193, COMPLETED 2023-05-16, no results posted (CT.gov NCT04943068)

Sources: PubMed 18206919 FDA NDA 210557 approval Vyleesi prescribing information PubMed 35147466 CT.gov NCT04943068

Tesamorelin [TH9507, GHRH(1-44) analogue]

It shrinks deep belly fat in adults who do not have HIV but carry excess weight around the middle.

Claim as graded “Reduces visceral (abdominal) fat in adults without HIV who carry excess belly fat”

Evidence tier C · provenance grade P4 · direction Supported · retail band 2, Moderate evidence. Derived by rule 8 - tier C with provenance P3+. Group: Prohormones, precursors and peptides. Scope qualifier: visceral fat only, in abdominally obese adults with reduced GH secretion; no effect on subcutaneous fat or BMI

Best available evidence: Stanley et al., "Metabolic effects of a growth hormone-releasing factor in obese subjects with reduced growth hormone secretion: a randomized controlled trial", J Clin Endocrinol Metab (PubMed 23015655), registered as NCT00675506 with results posted (CT.gov NCT00675506)

Sample and design: "A randomized, double-blind, placebo-controlled study" in "60 abdominally obese subjects with reduced GH secretion"; registry records allocation RANDOMIZED, QUADRUPLE masking (participant, care provider, investigator, outcomes assessor), Phase 2, actual enrolment 60, 29 vs 29 analysed; "Subjects received tesamorelin, a GHRH(1-44) analog, 2 mg once daily, or placebo for 12 months"; visceral adipose tissue by abdominal CT at L4, analysed by longitudinal linear mixed-effects modelling (PubMed 23015655; CT.gov NCT00675506)

Effect as reported: Visceral adipose tissue: "−16 ± 9 vs. 19 ± 9 cm(2), tesamorelin vs. placebo; treatment effect (95% confidence interval): −35 (−58, −12) cm(2); P = 0.003". Subcutaneous adipose tissue: "−6 ± 6 vs. 3 ± 11 cm(2); −10 (−32, +13) cm(2); P = 0.40" — i.e. no significant effect on subcutaneous fat. "No changes in fasting, 2-h glucose, or glycated hemoglobin were seen." (PubMed 23015655). In the approved HIV population, a 5-RCT meta-analysis found VAT MD = −27.71 cm², 95% CI [−38.37, −17.06], P < 0.001; trunk fat MD = −1.18 kg, 95% CI [−1.40, −0.96], P < 0.001; lean body mass MD = +1.42 kg, 95% CI [1.13, 1.71], P < 0.001; "no significant reductions in subcutaneous adipose tissue or BMI were observed" (Obes Res Clin Pract, PubMed 41545261)

Regulatory position: Approved — for a different indication than the claim. EGRIFTA (tesamorelin for injection) NDA 022505 was approved November 10, 2010 (FDA approval letter). The label indication is "for the reduction of excess abdominal fat in HIV-infected adult patients with lipodystrophy", and the Limitations of Use state "Long-term cardiovascular safety… has not been established" and, explicitly, that it is "Not indicated for weight loss management" (EGRIFTA SV label). Use in adults without HIV for belly fat is therefore outside the approved indication

Safety signal: Label warnings: "increased risk of neoplasms", elevated IGF-1, fluid retention, glucose intolerance and diabetes mellitus, hypersensitivity reactions, injection-site reactions, and "increased mortality in patients with acute critical illness"; contraindicated in disruption of the hypothalamic-pituitary axis, active malignancy, hypersensitivity and pregnancy; adverse reactions >5% arthralgia, injection site erythema and pruritus, pain in extremity, peripheral oedema, myalgia (EGRIFTA SV label). In the non-HIV obesity trial "There were no serious adverse events or differences in adverse events between the groups" (PubMed 23015655), with hypertension recorded in 8/31 tesamorelin vs 5/29 placebo and hyperglycaemia 3/31 vs 2/29 in the posted registry adverse-event table (CT.gov NCT00675506). Non-reporting: NCT00123253, the 412-participant Phase 3 pivotal trial in HIV-associated lipodystrophy, status COMPLETED with completion date 2007-04, no results posted (CT.gov NCT00123253). NCT01579695 (n = 391, long-term observational EGRIFTA study) and NCT01591902 (n = 129, diabetic retinopathy in EGRIFTA-treated HIV subjects) are both TERMINATED with no results posted (CT.gov API)

Sources: PubMed 23015655 CT.gov NCT00675506 PubMed 41545261 EGRIFTA SV label FDA NDA 022505 approval CT.gov NCT00123253

AOD-9604

It causes fat or weight loss in adults with obesity.

Claim as graded “Produces fat loss / weight loss in adults with obesity”

Evidence tier D · provenance grade P3 · direction Not supported · retail band 5, Evidence says no. Derived by rule 1 - direction Not supported. Group: Prohormones, precursors and peptides.

Best available evidence: FDA Pharmacy Compounding Advisory Committee evaluation of AOD-9604 (free base) and AOD-9604 acetate for obesity, which reviews the whole human dataset including the unpublished 536-subject "OPTIONS" trial (FDA briefing document)

Sample and design: Largest study: "A randomized (R), placebo-controlled (PC), DB trial, in adults (536 enrolled; 502 randomized) with obesity (18-65 years old, BMI 30-45 kg/m2)… oral AOD-9604 (0.25, 0.5, or 1 mg) or placebo once daily in addition to a dietician-supervised diet and exercise program for 24 weeks", powered for "an 80% chance of achieving significance on the primary endpoint if the weight loss compared to placebo was 1.8 kg". Supporting studies were far smaller: an IV dose-escalation study in 23 adults with obesity, an oral study in 16 men with obesity, a 1-week dose-escalation study in 36 men, and a 300-patient 12-week abstract-only trial (FDA briefing document)

Effect as reported: "Trial findings… indicate there was not a significant difference in the primary endpoint of weight loss between placebo and AOD-9604." Sponsor's own statement: "Trial results showed that weight loss compared to placebo at the primary and secondary endpoints of 12 or 24 weeks of treatment, was too low to reach statistical significance… the overall population did not respond consistently." Earlier IV study: "some weight loss occurred, on average 0.58 kg over 3 weeks, but this was not statistically different from placebo." Oral study: "there was no statistically significant weight loss compared with placebo." No confidence intervals or p-values are reported in any of the studies FDA could locate (FDA briefing document)

Regulatory position: Not an approved drug and recommended against compounding-list inclusion. FDA: "There is no applicable USP or NF drug substance monograph… and neither is a component of an FDA-approved drug"; "we believe the evaluation criteria weigh against placing AOD-9604 (free base) or AOD-9604 acetate on the [503A Bulks List]" (FDA briefing document; PCAC meeting minutes). FDA also records "A search of clinicaltrials.gov did not list any studies on AOD-9604", which the ClinicalTrials.gov API independently confirmed as 0 studies (FDA briefing document; CT.gov API)

Safety signal: No adverse-event signal has been captured, and FDA states plainly that this absence is not reassurance: "The FAERS search did not retrieve any reports, and the literature search did not identify any literature cases of adverse events"; "OSE added that the lack of reports for AOD-9604 does not imply that the substance is safe or lacks toxicities"; a CAERS search for 1/1/04 to 4/3/24 "did not retrieve cases"; "We did not find case reports on the use of AOD-9604 that include safety assessment"; and "FDA did not identify clinical studies in humans assessing pharmacokinetics or pharmacodynamics of AOD-9604 via any ROA". FDA further records that no human exposure data exist for the subcutaneous or transdermal routes — the routes it is actually sold for (FDA briefing document)

Sources: FDA PCAC briefing document FDA PCAC meeting minutes CT.gov API

Epitalon [epithalon, AEDG peptide, Ala-Glu-Asp-Gly]

It brings back natural melatonin production and gives older adults better sleep.

Claim as graded “Restores melatonin production and improves sleep in older adults ("anti-ageing" sleep claim)”

Evidence tier E · provenance grade P0 · direction Insufficient · retail band 6, Too little evidence yet. Derived by rule 3 - direction Insufficient. Group: Prohormones, precursors and peptides.

Best available evidence: FDA Pharmacy Compounding Advisory Committee evaluation of epitalon-related bulk drug substances for insomnia, dated 5/12/2026 (FDA briefing document)

Sample and design: n.a. for the claimed outcome. FDA: "After conducting a literature search, three studies were found in which epitalon was administered to humans", none of which studied insomnia. The largest, Ivko et al. 2021, enrolled 75 healthy women aged 40–50 primarily working night shifts, of whom 40 with low melatonin metabolite levels were split into a placebo group and an AEDG group of 20 each and treated sublingually for 20 days; FDA notes "The publication did not specify blinding, and additional limitations include short duration and small study size" and "The study did not evaluate sleep outcomes" (FDA briefing document). ClinicalTrials.gov returned 0 epitalon studies (CT.gov API)

Effect as reported: No numeric effect reported for sleep — the outcome was never measured. The only human outcome reported is a surrogate: 6-sulfatoxymelatonin "increased 1.7-fold after treatment", with the authors claiming "AEDG peptide restored the epiphyseal melatonin formation to a normal, age-appropriate value", and 83% vs 25% showing "distinct positive dynamics of psycho-emotional and general physical condition" — for which FDA records "no details were provided about how psycho-emotional and general physical condition were assessed". No confidence intervals or p-values are reported. FDA conclusion: "there is a lack of evidence to support the effectiveness of epitalon (free base) or epitalon acetate for the proposed use of insomnia, a disorder which increases the risk for serious conditions" (FDA briefing document)

Regulatory position: Not an approved drug. FDA: "There is no applicable United States Pharmacopeia (USP) or National Formulary (NF) drug substance monograph for epitalon (free base) or its acetate form, and neither is a component of an FDA-approved drug." The nominations for inclusion on the 503A compounding bulks list "were withdrawn and FDA is evaluating the substances at its discretion". An orphan drug designation for retinitis pigmentosa granted 2 September 2010 "was withdrawn/revoked on January 6, 2016" (FDA briefing document)

Safety signal: No safety dataset exists. FDA: "The Office of Surveillance and Epidemiology conducted a search of the FAERS database for reports of adverse events (AEs) for epitalon through December 8, 2025. The search retrieved no reports"; the Human Foods Complaint System search through December 5, 2025 "retrieved no cases"; the Ivko study "did not report safety data"; "The nominations did not include, and our search of the published medical literature has not identified clinical studies assessing the safety of epitalon-related BDSs administered in humans"; "No pharmacokinetic data were found" (FDA briefing document). Independent-replication problem: "It is critical to note that every preclinical and clinical study discussed here has been conducted by Dr. Khavinson's group in Russia with no independent confirmation of their results"; "Clinical studies on epithalon have not been conducted"; the 266-participant epithalamin mortality trial's "paper describing this trial is not available in English" (ADDF Cognitive Vitality report)

Sources: FDA PCAC epitalon briefing ADDF Cognitive Vitality report CT.gov API

Melatonin

It gets adults with insomnia to sleep faster, measured either in a sleep lab, by a wrist tracker, or by their own sleep diary.

Claim as graded “Reduces sleep onset latency in adults with insomnia (primary sleep disorders), measured by polysomnography/actigraphy or by sleep logs”

Evidence tier B · provenance grade P3 · direction Supported · retail band 1, Strong evidence. Derived by rule 6 - tier A/B with provenance P3+. Group: Recovery, sleep and adaptogens.

Best available evidence: Meta-analysis of randomised placebo-controlled trials (Ferracioli-Oda 2013, PLoS ONE); supporting meta-analysis in insomnia disorder (Maruani 2023, J Sleep Res)

Sample and design: 19 studies, 1683 subjects; 14 of 19 in insomnia. Second source: 22 studies, 4875 participants (925 on melatonin)

Effect as reported: Sleep onset latency WMD = 7.06 min, 95% CI 4.37 to 9.75, p<0.001 (fixed effects); objective measures only WMD 5.50 min, 95% CI 2.29 to 8.71; subjective measures only WMD 10.68 min, 95% CI 5.78 to 15.58. Heterogeneity I² = 56%, Q=31.9, p=0.004. PR-melatonin: objective SOL −5.05 min, p<0.001; subjective SOL −6.30 min, p=0.031 — no CIs reported

Regulatory position: EU authorised Art.13(1) health claim: "Melatonin contributes to the reduction of time taken to fall asleep" — conditions: "The claim may be used only for food which contains 1 mg of melatonin per quantified portion. In order to bear the claim, information shall be given to the consumer that the beneficial effect is obtained by consuming 1 mg of melatonin close to bedtime." (Reg. (EU) 432/2012, EFSA 2011;9(6):2241). Second authorised claim: "Melatonin contributes to the alleviation of subjective feelings of jet lag" (≥0.5 mg per portion). Also a prescription-only licensed medicine in the EU: Circadin 2 mg prolonged-release, "indicated as monotherapy for the short-term treatment of primary insomnia characterised by poor quality of sleep in patients who are aged 55 or over", marketing authorisation granted 29 June 2007, "can only be obtained with a prescription"

Safety signal: No serious signal found. EMA lists side effects in 1–10 per 1,000 including irritability, migraine, hypertension, hyperbilirubinaemia and abnormal liver function tests; "Side effects with Circadin are not common"

Sources: Ferracioli-Oda 2013, PLoS ONE Maruani 2023, J Sleep Res "Melatonin contributes to the reduction of time taken to fall asleep" "indicated as monotherapy for the short-term treatment of primary insomnia characterised by poor quality of sleep in patients who are aged 55 or over"

Magnesium

It gives adults who sleep badly better sleep, measured by how they rate it themselves.

Claim as graded “Improves sleep quality / insomnia severity in adults with poor sleep, measured by self-reported instruments (Insomnia Severity Index) or unspecified sleep-latency reporting”

Evidence tier D · provenance grade P1 · direction Contested · retail band 4, Experts disagree. Derived by rule 2 - direction Contested. Group: Recovery, sleep and adaptogens.

Best available evidence: Systematic review & meta-analysis of RCTs (Mah & Pitre 2021, BMC Complement Med Ther) plus a single large double-blind RCT (Schuster 2025, Nat Sci Sleep)

Sample and design: Pooled: 3 RCTs, 151 older adults. RCT: 155 randomised, 153 in ITT analysis, adults 18–65

Effect as reported: Pooled sleep onset latency 17.36 min less than placebo, 95% CI −27.27 to −7.44, p=0.0006; total sleep time +16.06 min, statistically insignificant. RCT: ISI change −3.9 (95% CI −5.8 to −2.0) vs placebo −2.3 (95% CI −4.1 to −0.4); between-group difference 1.6 (95% CI 0.0 to 3.3), p=0.049, Cohen's d = 0.20

Regulatory position: EU authorised claims for magnesium do not cover sleep. Authorised Art.13(1) wordings include "Magnesium contributes to a reduction of tiredness and fatigue", "Magnesium contributes to electrolyte balance", "Magnesium contributes to normal muscle function", "Magnesium contributes to normal psychological function" and "Magnesium contributes to normal functioning of the nervous system"; all carry the condition ["The claim may be used only for food which is at least a source of magnesium as referred to in the claim SOURCE OF [NAME OF VITAMIN/S] AND/OR [NAME OF MINERAL/S] as listed in the Annex to Regulation (EC) No 1924/2006."](https://ec.europa.eu/food/food-feed-portal/backend/claims/files/euregister.pdf) (Reg. (EU) 432/2012)

Safety signal: No serious signal found in fetched sources. In the 2025 RCT 7 adverse events on placebo vs 2 on magnesium; no serious adverse events; no discontinuations for adverse events

Sources: Mah & Pitre 2021, BMC Complement Med Ther Schuster 2025, Nat Sci Sleep "Magnesium contributes to a reduction of tiredness and fatigue"

Ashwagandha (Withania somnifera)

It lowers the stress hormone cortisol in the blood, and how stressed people say they feel, in adults under stress or with anxiety.

Claim as graded “Lowers serum/biological cortisol, and perceived stress, in adults with stress or anxiety”

Evidence tier B · provenance grade P2 · direction Contested · retail band 4, Experts disagree. Derived by rule 2 - direction Contested. Group: Recovery, sleep and adaptogens.

Best available evidence: Systematic review & meta-analysis of RCTs (Albalawi 2025, Nutr Health); second meta-analysis (Bachour 2025, BJPsych Open)

Sample and design: 7 studies for cortisol, 6 for perceived stress; n = 488 participants. Second source: 15 studies, 873 patients

Effect as reported: Cortisol −1.16 µg/dL, 95% CI −1.64 to −0.69, P<0.001; perceived stress (PSS) SMD = −0.355, 95% CI −1.188 to 0.47, P = 0.40 — not significant; heterogeneity I² = 50.9%. Second meta-analysis reports cortisol μ = −2.3626, 95% CI −3.2622 to −1.4629, p<0.0001 and PSS μ = −4.88, 95% CI −7.84 to −1.91, p=0.0013 at 8 weeks

Regulatory position: Ashwagandha: no entry located in the EU Register of nutrition and health claims — i.e. no authorised health claim; not evaluated in that register. No EMA herbal monograph fetched. WADA status not checked (WADA list not fetched)

Safety signal: Yes — hepatotoxicity. LiverTox: ashwagandha "has not been implicated in causing serum enzyme elevations during therapy but recently has been implicated in cases of clinically apparent liver injury", likelihood score B = "Likely cause of clinically apparent liver injury"; onset 2–12 weeks, usually cholestatic or mixed with jaundice and pruritus; fatal liver injury, acute liver failure, acute-on-chronic liver failure and emergent liver transplantation reported, particularly in pre-existing liver disease and cirrhosis; use "should be avoided in patients with cirrhosis or advanced chronic liver disease"

Sources: Albalawi 2025, Nutr Health Bachour 2025, BJPsych Open no entry located in the EU Register of nutrition and health claims "has not been implicated in causing serum enzyme elevations during therapy but recently has been implicated in cases of clinically apparent liver injury"

Rhodiola rosea

It cuts physical or mental tiredness in adults.

Claim as graded “Reduces physical or mental fatigue in adults”

Evidence tier D · provenance grade P1 · direction Insufficient · retail band 6, Too little evidence yet. Derived by rule 3 - direction Insufficient. Group: Recovery, sleep and adaptogens.

Best available evidence: Systematic review of RCTs/CCTs, no meta-analysis possible (Ishaque 2012, BMC Complement Altern Med)

Sample and design: 11 studies (10 RCTs, 1 CCT), 446 subjects; 6 physical-performance studies, 5 mental-fatigue studies

Effect as reported: `no numeric effect reported` — meta-analysis was not performed because no two studies reported the same outcomes; 2 of 6 physical-fatigue trials and 3 of 5 mental-fatigue RCTs reported R. rosea effective

Regulatory position: EU: traditional-use herbal medicinal product. EMA HMPC European Union herbal monograph for Rhodiolae roseae rhizoma et radix: "On the basis of its long-standing use, arctic root can be used for the temporary relief of symptoms of stress, such as fatigue and sensation of weakness"; adults over 18 only; not longer than two weeks without medical advice; the HMPC conclusion is explicitly based on traditional use because "there is insufficient evidence from clinical trials". No authorised Art.13(1) food health claim for Rhodiola located in the EU Register

Safety signal: No serious signal found. EMA: "Arctic root medicines are generally well tolerated" and "At the time of the HMPC assessment, no side effects had been reported with these medicines". One trial participant had a minor headache and another insomnia during long-term supplementation (Ishaque 2012)

Sources: Ishaque 2012, BMC Complement Altern Med "On the basis of its long-standing use, arctic root can be used for the temporary relief of symptoms of stress, such as fatigue and sensation of weakness" EU Register

Omega-3 EPA/DHA

It reduces the muscle soreness that shows up two days after hard, damaging exercise.

Claim as graded “Reduces delayed onset muscle soreness two days after eccentric exercise in adults”

Evidence tier B · provenance grade P1 · direction Contested · retail band 4, Experts disagree. Derived by rule 2 - direction Contested. Group: Recovery, sleep and adaptogens.

Best available evidence: Systematic review & meta-analysis of RCTs, PROSPERO CRD42018085869 (Lv 2020, Biomed Res Int)

Sample and design: 12 RCTs; 145 subjects and 156 controls

Effect as reported: DOMS at 48 h MD = −0.93, 95% CI −1.44 to −0.42, p=0.0004 on a 10-unit VAS — below the stated minimal clinically important difference of 1.4; isometric muscle strength and range of motion: no significant differences

Regulatory position: No authorised EU claim for recovery or soreness. The EU Register contains authorised EPA/DHA claims only for "the normal function of the heart", and DHA claims for brain function and vision; no authorised EPA/DHA claim relating to recovery, muscle soreness, muscle fatigue or exercise recovery. Joint-related wordings are non-authorised, e.g. "Omega-3 EPA and DHA help maintain healthy joints", reason: "this claimed effect for this food has not been substantiated" (EFSA 2009;7(9):1263)

Safety signal: No serious signal found in fetched sources

Sources: Lv 2020, Biomed Res Int authorised EPA/DHA claims only for "the normal function of the heart", and DHA claims for brain function and vision; no authorised EPA/DHA claim relating to recovery, muscle soreness, muscle fatigue or exercise recovery

Vitamin D

It makes adults stronger overall — but only those starting with genuinely low vitamin D in their blood.

Claim as graded “Increases global muscle strength in adults with low baseline 25(OH)D (<30 nmol/L)”

Evidence tier B · provenance grade P3 · direction Contested · retail band 4, Experts disagree. Derived by rule 2 - direction Contested. Group: Recovery, sleep and adaptogens.

Best available evidence: Systematic review & meta-analysis of RCTs (Beaudart 2014, JCEM); contrasting newer meta-analysis (2024, 35 RCTs)

Sample and design: 30 RCTs, 5615 participants; primary outcome from 29 RCTs, 5533 participants; deficient subgroup 9 studies, 710 participants

Effect as reported: Global muscle strength overall SMD 0.17, 95% CI 0.03 to 0.31, P=.02, heterogeneity I² 77.67%; baseline 25(OH)D <30 nmol/L SMD 0.47, 95% CI −0.07 to 1.01, P=.02 for the between-subgroup comparison vs ≥30 nmol/L SMD 0.06, 95% CI −0.05 to 0.16; grip strength alone SMD 0.01, 95% CI −0.06 to 0.07, P=.87; lower-limb strength SMD 0.19, 95% CI 0.05 to 0.34, P=.01. Newer pooled handgrip result SMD 0.08, 95% CI −0.06 to 0.21, p=.26, I²=65%, high-dose subgroup SMD 0.31, 95% CI 0.07 to 0.55, p=.01

Regulatory position: EU authorised Art.13(1) claim: "Vitamin D contributes to the maintenance of normal muscle function" — condition: ["The claim may be used only for food which is at least a source of vitamin D as referred to in the claim SOURCE OF [NAME OF VITAMIN/S] AND/OR [NAME OF MINERAL/S] as listed in the Annex to Regulation (EC) No 1924/2006."](https://ec.europa.eu/food/food-feed-portal/backend/claims/files/euregister.pdf) (Reg. (EU) 432/2012, EFSA 2010;8(2):1468). Other authorised claims include bones, teeth, blood calcium, immune function and cell division (EU Register)

Safety signal: Yes — hypervitaminosis D. Vitamin D toxicity leads to hypercalcemia with serum calcium often >11 mg/dL and 25(OH)D often >150 ng/mL; manifestations include nephrolithiasis, cardiac arrhythmias, pancreatitis, stupor and coma; severe hypercalcemia can lead to acute renal failure requiring hemodialysis. 25,397 cases of vitamin D toxicity were reported from 2000 to 2014

Sources: Beaudart 2014, JCEM 2024, 35 RCTs "Vitamin D contributes to the maintenance of normal muscle function" Vitamin D toxicity leads to hypercalcemia

Zinc

It raises testosterone in the blood of adult men.

Claim as graded “Raises serum total testosterone in adult men”

Evidence tier D · provenance grade P3 · direction Contested · retail band 4, Experts disagree. Derived by rule 2 - direction Contested. Group: Recovery, sleep and adaptogens.

Best available evidence: Systematic review, no pooled estimate (Te Liu / J Trace Elem Med Biol 2023); narrative review of human intervention studies (Santos & Teixeira 2020, Aging Male)

Sample and design: 38 papers: 8 clinical and 30 animal studies; total human participants not stated. Narrative review: 1070 titles screened, 12 full-length publications included; total participants not stated

Effect as reported: `no numeric effect reported` for any pooled estimate — no pooled effect size, no 95% CI, no meta-analytic summary statistic is reported. Individual trials only, no CIs: in 100 haemodialysis men, total testosterone increased by 400 ng/dL after 6 weeks of zinc sulfate 250 mg/day; chelated zinc 30 mg/day for 1–6 months total testosterone increased from 180 to 222 ng/dL, while chelated zinc 30 mg/day for 8 weeks produced no increase in testosterone; in idiopathic hypogonadotropic hypogonadism, adding zinc to hCG+FSH gave similar efficacy to hCG+FSH alone

Regulatory position: EU authorised Art.13(1) claim: "Zinc contributes to the maintenance of normal testosterone levels in the blood" — condition: ["The claim may be used only for food which is at least a source of zinc as referred to in the claim SOURCE OF [NAME OF VITAMIN/S] AND/OR [NAME OF MINERAL/S] as listed in the Annex to Regulation (EC) No 1924/2006."](https://ec.europa.eu/food/food-feed-portal/backend/claims/files/euregister.pdf) (Reg. (EU) 432/2012, EFSA 2010;8(10):1819, entry Id 301). Note the authorised wording is maintenance of normal levels, not elevation above normal

Safety signal: Yes — copper deficiency myelopathy from chronic zinc excess. Documented case: 62-year-old man with copper deficiency myeloneuropathy secondary to chronic use of a zinc-containing dental fixative; serum copper 6.4 µmol/L (ref 10–22), serum zinc 20.3 µmol/L (ref 11–18); after treatment he remained largely wheelchair-dependent and did not regain the ability to walk independently. Also gastric discomfort at 660 mg/day zinc sulfate requiring dose reduction (Netter et al., in Santos & Teixeira)

Sources: Te Liu / J Trace Elem Med Biol 2023 Santos & Teixeira 2020, Aging Male "Zinc contributes to the maintenance of normal testosterone levels in the blood" copper deficiency myeloneuropathy secondary to chronic use of a zinc-containing dental fixative

Collagen peptides

It eases tendon or joint pain and gets joints moving better in active adults.

Claim as graded “Reduces tendon or joint pain and improves joint function in active adults”

Evidence tier C · provenance grade P1 · direction Supported · retail band 3, Widely studied, poorly measured. Derived by rule 5 - provenance cap at P1 or below. Group: Recovery, sleep and adaptogens.

Best available evidence: Systematic review of RCTs, no meta-analysis performed (Khatri 2021, Amino Acids); strongest single trial is a double-blind placebo-controlled cross-over RCT in Achilles tendinopathy (Praet 2019, Nutrients)

Sample and design: Review: 15 RCTs, 656 participants (325 male, 276 female), 12 studies in recreational athletes; all double-blinded and placebo-controlled or isonitrogenous-controlled; no pooled quantitative estimate is reported. Trial: 20 patients, cross-over, ANZCTR#12615001035516

Effect as reported: No pooled effect exists. Trial VISA-A improvement over first 3 months: collagen-first group 12.6 points, 95% CI 9.7–15.5 vs placebo-first group 5.3 points, 95% CI 2.3–8.3, against a stated MCID of 6.5 points; after crossover 5.9 points (95% CI 2.8–9.0) and 17.7 points (95% CI 14.6–20.7). Review-reported single-study effect sizes: knee joint pain when walking ES = 0.36; pain during activity Cohen's d = 0.30

Regulatory position: Non-authorised in the EU. "Collagen can/could contribute to the maintenance of the healthy function of joints" — Non-authorised, reason: "this claimed effect for this food has not been substantiated" (EFSA 2011;9(6):2247). Collagen hydrolysate wordings including "Contributes to improved joint functioning and joint mobility" and "Characteristic collagen peptide mixture (collagen hydrolysate) having a beneficial physiological effect on the maintenance of joint health in physically active people" are also Non-authorised (Reg. (EU) 379/2012)

Safety signal: No serious signal found in fetched sources. Note funding: the Achilles trial was financially supported by GELITA AG, Germany, the manufacturer of the tested product

Sources: Khatri 2021, Amino Acids Praet 2019, Nutrients "Collagen can/could contribute to the maintenance of the healthy function of joints"

Tart cherry

It gets muscle strength back faster after exercise that damaged it, in athletes.

Claim as graded “Accelerates recovery of muscle function (maximal voluntary contraction) after muscle-damaging exercise in athletes”

Evidence tier B · provenance grade P2 · direction Supported · retail band 2, Moderate evidence. Derived by rule 7 - tier A/B with provenance P2. Group: Recovery, sleep and adaptogens.

Best available evidence: Systematic review & meta-analysis, PRISMA 2020, OSF-registered (Daab 2026, Sports Med Open); earlier meta-analysis (Hill 2021, IJSNEM)

Sample and design: 19 trials, 385 participants; 8 crossover and 11 parallel; double-blinding in 11 studies, single-blinding in 3. Earlier meta-analysis: 14 studies, total participants not stated

Effect as reported: MVC recovery: post-exercise ES 0.63, 95% CI 0.05 to 1.22, I²=69.30%, p=0.034; 24 h ES 1.12, 95% CI 0.20 to 2.05, I²=91.15%; 48 h ES 1.29, 95% CI 0.25 to 2.34, I²=92.67%; 72 h ES 2.14, 95% CI 0.34 to 3.94, I²=92.38%. Muscle soreness was null at every time point, e.g. 48 h ES −0.39, 95% CI −1.14 to 0.36, p=0.305; CK null at every time point, e.g. 48 h ES −0.08, 95% CI −0.51 to 0.35. Earlier meta-analysis reported strength ES −0.78, 95% CI −1.11 to −0.46 and soreness ES −0.44, 95% CI −0.87 to −0.02

Regulatory position: Non-authorised in the EU. ["[Tart/sour] cherries help support healthy joints"](https://ec.europa.eu/food/food-feed-portal/backend/claims/files/euregister.pdf) — Non-authorised, reason: "this claimed effect for this food has not been substantiated" (EFSA 2010;8(2):1493). ["[Tart/sour] cherries provide a rich source of antioxidants"](https://ec.europa.eu/food/food-feed-portal/backend/claims/files/euregister.pdf) — Non-authorised, reason: "this claimed effect for this food is not a beneficial physiological effect as required by the Regulation". No authorised recovery claim located

Safety signal: No serious signal found in fetched sources

Sources: Daab 2026, Sports Med Open Hill 2021, IJSNEM "this claimed effect for this food has not been substantiated"

Curcumin (from turmeric, Curcuma longa)

After exercise it lowers a blood marker of muscle damage, and the soreness that goes with it.

Claim as graded “Reduces exercise-induced muscle damage (creatine kinase) and muscle soreness in adults after exercise”

Evidence tier C · provenance grade P2 · direction Contested · retail band 4, Experts disagree. Derived by rule 2 - direction Contested. Group: Recovery, sleep and adaptogens.

Best available evidence: Two systematic reviews & meta-analyses of RCTs (Fang 2024, PLoS One; Fernández-Lázaro 2020, J Diet Suppl)

Sample and design: 14 articles, 349 subjects; 12 independent samples pooled for CK. Second meta-analysis: total participants and number of studies not stated on the fetched page

Effect as reported: CK MD = −137.32, 95% CI −238.82 to −35.82, Z=2.65, p=0.008, I² = 97%; the effect was confined to subgroups — untrained MD −282.55, 95% CI −455.29 to −109.82, p=0.001 vs trained MD 11.22, 95% CI −17.59 to 40.03, p=0.45 — not significant; no significant CK effect at any individual post-exercise time point (24 h 8.32, 95% CI −15.51 to 32.16; 48 h 3.93, 95% CI −127.99 to 135.84; 72 h −18.00, 95% CI −100.53 to 64.54). Second meta-analysis: CK WMD = −48.54 IU·L⁻¹, 95% CI −80.667 to −16.420, p=.003 and muscle soreness index WMD = −0.476, 95% CI −0.750 to −0.202, p=.001

Regulatory position: Curcumin: no entry located in the EU Register of nutrition and health claims — no authorised health claim for exercise recovery or inflammation

Safety signal: Yes — hepatotoxicity. LiverTox likelihood score A: turmeric/curcumin "recently established as a well-documented cause of clinically apparent liver injury", with "several dozen instances of clinically apparent acute liver injury"; predominantly hepatocellular with ALT often >1,000 U/L; some cases severe and unremitting, leading to acute liver failure, death, or need for liver transplantation; most cases attributed to highly bioavailable curcumin forms using piperine/black pepper or lipid nanoparticle delivery; HLA-B*35:01 found in over 70% of cases vs 10–15% of controls; turmeric "appears to have become the most common cause of clinically apparent, herbal-related liver injury in the United States"

Sources: Fang 2024, PLoS One Fernández-Lázaro 2020, J Diet Suppl no entry located in the EU Register of nutrition and health claims "recently established as a well-documented cause of clinically apparent liver injury"

Enobosarm (ostarine, MK-2866, GTx-024)

It adds muscle, measured on a body scan, in adults — healthy older people, people wasting from cancer, and people on GLP-1 drugs.

Claim as graded “Increases total lean body mass, measured by DXA, in adults (healthy elderly, cancer-cachexia and GLP-1-treated populations)”

Evidence tier C · provenance grade P4 · direction Supported · retail band 2, Moderate evidence. Derived by rule 8 - tier C with provenance P3+. Group: SARMs and research peptides.

Best available evidence: Two identically designed randomised, double-blind, placebo-controlled phase 3 trials (POWER 1 / POWER 2), plus multiple phase 2 RCTs. No pooled meta-analysis found. (POWER 1/2, ASCO abstract; NCT01355484; phase II, Dalton 2011)

Sample and design: No pooled source exists. Largest individual trials: POWER 1 n=321, POWER 2 n=320 (ASCO); phase 2 in cancer n=159 (LiverTox); phase 2 in healthy elderly n=120, 115 completers (PMC3177038); phase 2b QUALITY n=168 (Veru)

Effect as reported: POWER 1 day 84 LBM +0.41 kg vs placebo −0.92 kg, p=0.0002; day 147 p<0.0001. POWER 2 day 84 +0.47 kg vs −0.37 kg, p=0.0111; day 147 p=0.0028 (ASCO). Phase 2, 3 mg administered daily for 86 days: +1.3 kg ± 0.3 (±1 SE) vs placebo, p<0.001; 1 mg +0.7 kg ± 0.3, p=0.055 (PMC3177038). No 95% CIs reported in any fetched source.

Regulatory position: Not an authorised medicine anywhere; EDQM states these performance-enhancing substances "have not been approved for medical use, since clinical studies performed up till now have not demonstrated their safety/efficacy" (EDQM MSSIP006.pdf/c816b3f9-11ab-a2bd-39ca-f4699440ecb1?t=1736258491590)). WADA 2026 Prohibited List S1.2 Other Anabolic Agents, naming "enobosarm (ostarine)", prohibited at all times (WADA 2026 List)

Safety signal: Hepatotoxicity. In the phase 2 healthy-elderly trial ALT elevations occurred in 21% at the highest dose vs none on placebo, and 1 subject on 3 mg daily discontinued with ALT 4.2× ULN (LiverTox; discontinuation also in PMC3177038). Ostarine appears among published SARM drug-induced liver injury cases (Koller 2021; Cureus review). FDA warns SARM-containing bodybuilding products have caused "life-threatening reactions, including liver injuries that required hospitalization" (FDA)

Sources: POWER 1/2, ASCO abstract NCT01355484 phase II, Dalton 2011 LiverTox Veru EDQM MSSIP006 WADA 2026 List Koller 2021 Cureus review FDA

Ligandrol (LGD-4033, VK5211)

Taken for three weeks, it adds muscle in healthy men.

Claim as graded “Increases lean body mass in healthy men over short-term (21-day) administration”

Evidence tier D · provenance grade P4 · direction Supported · retail band 6, Too little evidence yet. Derived by rule 4 - tier D or E. Group: SARMs and research peptides. Scope qualifier: lean mass only; strength not supported

Best available evidence: A single phase 1 randomised, double-blind, placebo-controlled ascending-dose trial (Basaria et al.). No second efficacy RCT and no pooled source found (PMC4111291)

Sample and design: n=76 healthy men aged 21–50 randomised (33 placebo, 18 × 0.1 mg, 11 × 0.3 mg, 14 × 1.0 mg); 68 completers; DXA evaluable 30/17/10/11. 1 study. 21 days of treatment (PMC4111291)

Effect as reported: LBM increased dose-dependently, p for trend = 0.04; increase averaged 1.21 kg at the 1.0-mg dose administered daily (p = 0.047 vs placebo). No 95% CI reported. Appendicular skeletal muscle mass change was not significant (p for trend = 0.078). Strength increase averaged 68.3 N at 1.0 mg but was not significantly different from placebo; stair-climbing speed and power showed a non-significant trend (PMC4111291)

Regulatory position: Not an authorised medicine; captured by the EDQM finding above for SARMs (EDQM MSSIP006.pdf/c816b3f9-11ab-a2bd-39ca-f4699440ecb1?t=1736258491590)). WADA 2026 List S1.2, naming "LGD-4033 (ligandrol)" (WADA)

Safety signal: Testosterone suppression and hepatotoxicity. Dose-dependent suppression of total testosterone, SHBG, HDL cholesterol and triglycerides; FSH suppressed at 1.0 mg; hormones returned to baseline by day 56 (PMC4111291). The trial itself found "no clinically significant changes in liver enzymes" (LiverTox), but cholestatic drug-induced liver injury is documented outside trials: a 19-year-old athlete who took one ligandrol capsule daily for 4 weeks developed jaundice with canalicular cholestasis and ductopenia on biopsy (Koller 2021, World J Clin Cases); ligandrol accounts for 4 of 6 published SARM-DILI cases reviewed (Cureus review)

Sources: PMC4111291 EDQM MSSIP006 WADA LiverTox Koller 2021, World J Clin Cases Cureus review

RAD-140 (testolone; developed as vosilasarm)

It adds muscle or strength in people.

Claim as graded “Increases lean body mass or muscle strength in humans”

Evidence tier E · provenance grade P0 · direction Insufficient · retail band 6, Too little evidence yet. Derived by rule 3 - direction Insufficient. Group: SARMs and research peptides.

Best available evidence: None for this claim. The only registered first-in-human study is a phase 1 oncology dose-escalation trial in metastatic breast cancer whose endpoints were dose-limiting toxicity, adverse events, pharmacokinetics and tumour response — not lean mass or strength (NCT03088527; phase 1 as summarised)

Sample and design: n.a. for the stated claim. The phase 1 trial enrolled 22 postmenopausal women with stage IV ER+/HER2− breast cancer (PMC10054042); no participants have been studied for lean mass or strength in any source fetched

Effect as reported: no numeric effect reported for lean mass or strength in humans

Regulatory position: Not an authorised medicine; captured by the EDQM SARM finding (EDQM MSSIP006.pdf/c816b3f9-11ab-a2bd-39ca-f4699440ecb1?t=1736258491590)). WADA 2026 List S1.2, naming "RAD140" (WADA)

Safety signal: Severe drug-induced liver injury, multiple cases. In the phase 1 trial the most common adverse effects were elevated AST 59.1%, ALT 45.5%, total bilirubin 27.3%, hypophosphatemia 22.7%, with more hyperbilirubinemia and hypophosphatemia at the 150 mg daily dose group (PMC10054042). Case: 26-year-old man, jaundice and acute liver injury, ALT 243 IU/L, total bilirubin 4.9 mg/dL, 4-day hospitalisation, normalised 1 month after cessation (J Med Case Rep 2023). Case: 22-year-old man after 16 weeks of RAD-140, total bilirubin peaked at 530 µmol/L, direct bilirubin 294 µmol/L, ALP 5.3 µkat/L, biopsy showing cholestasis (Cureus 2024). Published SARM-DILI cases took 3 to 12 months for liver enzymes to normalise (Cureus 2024)

Sources: NCT03088527 phase 1 as summarised EDQM MSSIP006 WADA Cureus 2024

Cardarine (GW501516, GW1516, endurobol)

It lets people go longer and harder.

Claim as graded “Increases endurance / exercise capacity in humans”

Evidence tier E · provenance grade P0 · direction Insufficient · retail band 6, Too little evidence yet. Derived by rule 3 - direction Insufficient. Group: SARMs and research peptides.

Best available evidence: None for this claim. Human trials of GW501516 that were fetched measured lipids only, in hospitalised sedentary volunteers explicitly prohibited from rigorous physical activity (Sprecher 2007, ATVB). Endurance data are rodent only (PMC6475847; Wikipedia summary of the 2007 Cell mouse study)

Sample and design: n.a. for endurance. The fetched human study randomised n=24 healthy normolipidaemic men (placebo n=6; 2.5 mg n=9; 10 mg n=9) to once-daily oral dosing for 14 days, for lipid endpoints (ATVB)

Effect as reported: no numeric effect reported for endurance or exercise capacity in humans

Regulatory position: Development terminated; "Clinical approval has not, and will not be given for this substance" (WADA alert). Not approved for medical use in Europe (EDQM MSSIP006.pdf/c816b3f9-11ab-a2bd-39ca-f4699440ecb1?t=1736258491590), which lists cardarine as a metabolic modulator in scope). WADA 2026 List S4.4.1 Metabolic Modulators, naming "(GW1516, GW501516)" as a PPARδ agonist (WADA)

Safety signal: CARCINOGENICITY IN TWO SPECIES. GSK abstracts report a rat carcinogenicity study and a mouse carcinogenicity study in which GW501516 caused cancer in rats and mice after 104 weeks of dosing; neither has been published as a full peer-reviewed paper (Mitchell & Bishop-Bailey, Pulmonary Circulation). WADA states the drug "was withdrawn from research by the pharmaceutical company and terminated when serious toxicities were discovered in pre-clinical studies", and took "the rare step of warning 'cheats'" because "the side effect of this chemical compound is so serious" (WADA alert). A secondary encyclopaedic source records the cancer as arising "rapidly in several organs, at dosages of 3 mg/kg/day in both mice and rats" (Wikipedia) — recorded here as a secondary attribution, not primary. Human carcinogenicity is not confirmed in humans (PMC6475847)

Sources: Sprecher 2007, ATVB PMC6475847 Wikipedia summary of the 2007 *Cell* mouse study WADA alert EDQM MSSIP006 WADA

MK-677 (ibutamoren, ibutamoren mesylate, MK-0677)

It raises growth hormone and a growth-related blood hormone, and adds muscle, in healthy older adults.

Claim as graded “Raises GH/IGF-1 and increases fat-free mass in healthy older adults”

Evidence tier D · provenance grade P4 · direction Supported · retail band 6, Too little evidence yet. Derived by rule 4 - tier D or E. Group: SARMs and research peptides. Scope qualifier: for GH/IGF-1 and fat-free mass; Not supported for strength or physical function

Best available evidence: A 2-year double-blind, randomised, placebo-controlled modified-crossover RCT (Nass et al.), powered for the 12-month endpoint (PMC2757071)

Sample and design: n=71 randomised; 65 completed year 1 (23 men, 42 women, ages 60–81; placebo n=22, MK-677 n=43); 59 completed 18 months, 53 completed 24 months. 1 pivotal study; no pooled source found (PMC2757071)

Effect as reported: At 12 months on 25 mg administered orally once daily: FFM (DXA) +1.1 kg (95% CI 0.7 to 1.5) with MK-677 vs −0.5 kg (95% CI −1.1 to 0.2) with placebo, p<0.001; 4-compartment model +1.3 kg (0.7 to 1.8) vs −0.4 kg (−1.1 to 0.4), p=0.001. Serum IGF-1 rose 1.5-fold (95% CI 1.4 to 1.6), p<0.001. However: "Increased FFM did not result in changes in strength or function"; no significant change in knee extension/flexion or shoulder extension, and "no significant changes in any measurements of function" (PMC2757071)

Regulatory position: Not an authorised medicine; ibutamoren is named by EDQM among small-molecule growth-hormone secretagogues sold as performance enhancers and "not approved for medical use" (EDQM MSSIP006.pdf/c816b3f9-11ab-a2bd-39ca-f4699440ecb1?t=1736258491590)). WADA 2026 List S2.2.4 Growth hormone releasing factors, naming "ibutamoren (MK-677)" as a growth hormone secretagogue (WADA)

Safety signal: Glucose intolerance and joint pain leading to dose reduction / withdrawal. In the RCT, study drug was back-titrated to 10 mg daily in 4 subjects: increased fasting glucose in an 81-year-old man and in 1 woman (withdrawn after 3 months), and increased joint pain in 2 women (both withdrew at 12 months) (PMC2757071). FDA's general assessment of growth-hormone secretagogues notes "known potential risks associated with elevated" GH/IGF-1 and, for this drug class, effects such as glucose intolerance and diabetes mellitus (FDA PCAC ipamorelin briefing)

Sources: PMC2757071 EDQM MSSIP006 WADA FDA PCAC ipamorelin briefing

BPC-157 (Body Protection Compound 157, PL-14736, bepecin)

It speeds up tendon and soft-tissue healing in people.

Claim as graded “Accelerates tendon or soft-tissue healing in humans”

Evidence tier E · provenance grade P0 · direction Insufficient · retail band 6, Too little evidence yet. Derived by rule 3 - direction Insufficient. Group: SARMs and research peptides.

Best available evidence: No human efficacy trial for this claim. A 2025 systematic review of 544 records included 36 studies: 35 preclinical, 1 clinical — and the single clinical study was a retrospective knee-pain series, not a tendon or soft-tissue healing study (Vasireddi et al., HSS Journal). A 2025 narrative review states "Only three pilot studies have examined BPC-157 in humans" — intra-articular knee pain, interstitial cystitis, and IV safety/pharmacokinetics (PMC12446177)

Sample and design: n.a. for tendon/soft-tissue healing. All human exposure to date, across the fetched reviews: 16 patients (retrospective knee injections), 12 (interstitial cystitis pilot), 2 (IV safety pilot, infusions up to 20 mg administered) (PMC12446177). A phase 2 randomised, double-blind, placebo-controlled trial in acute grade II hamstring strain is registered as NCT07437547 with MRI-assessed injury volume and return-to-sport co-primary endpoints; no results are posted (NCT07437547)

Effect as reported: no numeric effect reported for tendon or soft-tissue healing. The nearest human datum is uncontrolled: 14 of 16 patients reported "significant pain relief" 6 months to 1 year after knee injections, with "small sample size, lack of control group, lack of common diagnosis" and no p-value or CI (PMC12446177); the systematic review reports the same series as "7 of 12 patients" with relief >6 months (HSS Journal)

Regulatory position: Not an approved product anywhere: "There is no approved product containing BPC-157-related BDSs in any country at this time." No monograph in the European Pharmacopoeia (11th Ed., 11.8), Japanese Pharmacopoeia or International Pharmacopoeia. FDA proposes not adding BPC-157 free base or acetate to the 503A Bulks List, citing "lack of evidence of effectiveness" (FDA PCAC briefing, July 2026). WADA 2026 List S0 Non-Approved Substances, which "covers many different substances including but not limited to BPC-157" (WADA)

Safety signal: No clinical safety data. The systematic review states plainly: "No clinical safety data were found", and flags "potential risks due to unregulated production and lack of clinical safety data" (HSS Journal). FDA judged both BPC-157 substances "not well-characterized from the physical and chemical characterization perspective" and concluded that characterization, historical use, effectiveness and safety information together "weigh against them being added to the 503A Bulks List" (FDA). No serious adverse events were reported in the three small human pilots (PMC12446177)

Sources: Vasireddi et al., *HSS Journal* PMC12446177 NCT07437547 FDA PCAC briefing, July 2026 WADA

TB-500 / thymosin β4 (Tβ4; pharmaceutical designations RGN-352 injectable, RGN-259 ophthalmic)

It repairs damaged muscle, bone or heart tissue in people.

Claim as graded “Promotes tissue repair (musculoskeletal or cardiac) in humans”

Evidence tier E · provenance grade P0 · direction Insufficient · retail band 6, Too little evidence yet. Derived by rule 3 - direction Insufficient. Group: SARMs and research peptides. Scope qualifier: for tissue repair); the one completed randomised endpoint fetched was Not supported

Best available evidence: No completed human trial demonstrating tissue repair. The best-designed human trial fetched is a randomised, double-masked, placebo-controlled phase 2 ophthalmic trial that missed both co-primary endpoints (Sosne et al., Clin Ophthalmol). A cardiac trial of injectable Tβ4 after acute myocardial infarction is registered but no results are posted (NCT01311518)

Sample and design: n=72 randomised 1:1 to 0.1% Tβ4 or placebo for 28 days in the dry-eye phase 2 trial (PMC4445951). For musculoskeletal tissue repair: n.a. — no human trial was found in this session. A further registered study, NCT07487363 (TB-500 fragment, cardiovascular biomarkers in stable ASCVD, sponsor Hudson Biotech), shows "No Results" posted (NCT07487363)

Effect as reported: Dry-eye phase 2: "Neither of the primary endpoints, ie, ocular discomfort or inferior corneal staining, showed a significant difference between treatment and control groups." Inferior corneal staining 2.08 (Tβ4) vs 1.90 (placebo), difference 0.18, p=0.2586; ocular discomfort 1.6 vs 1.3, difference 0.3, p=0.2210 (PMC4445951). For tissue repair as claimed: no numeric effect reported

Regulatory position: Not an authorised medicine in the EU or elsewhere in any fetched source; n.a. for a specific EU legal determination (searched EMA and EDQM; the EDQM MSSIP006 study explicitly placed "illegal peptides" outside its scope). WADA 2026 List S2.3 Growth Factors and Growth Factor Modulators, naming "Thymosin-ß4 and its derivatives e.g. TB-500" (WADA)

Safety signal: No serious harm signal identified in the fetched sources. The dry-eye trial reported no safety concern in the sections fetched; the cardiac trial has posted no adverse-event data (NCT01311518). Absence of published safety data for systemic musculoskeletal use is itself the finding

Sources: Sosne et al., *Clin Ophthalmol* NCT01311518 NCT07487363 WADA

CJC-1295 (with DAC) combined with ipamorelin

Used together, they raise growth hormone in adults.

Claim as graded “Increases growth hormone secretion in adults”

Evidence tier E · provenance grade P0 · direction Insufficient · retail band 6, Too little evidence yet. Derived by rule 3 - direction Insufficient. Group: SARMs and research peptides. Scope qualifier: FOR THE COMBINATION; THE SINGLE-AGENT CJC-1295 GH/IGF-1 DATA WOULD BE P3 for the combination); CJC-1295 alone raises GH/IGF-1

Best available evidence: No human study of the combination. Human evidence exists only for each agent separately: for CJC-1295, "two randomized, placebo-controlled, double-blind, ascending dose trials with durations of 28 and 49 d" in healthy adults (Teichman et al. 2006, JCEM); for ipamorelin, a phase 2 RCT in postoperative ileus (Beck et al. 2014)

Sample and design: CJC-1295 alone: healthy subjects aged 21–61; the fetched abstract does not state the number enrolled (`n.a.`) (JCEM); a separate FDA presentation records "6 RCTs reported" for CJC-1295 and a 12-week phase II in 192 HIV patients with lipodystrophy (FDA presentation). Ipamorelin alone: n=117 enrolled, 114 in the safety and mITT populations (Beck 2014, NCT00672074). Combination: 0 human participants in any source fetched

Effect as reported: CJC-1295 alone: "dose-dependent increases in mean plasma GH concentrations by 2- to 10-fold for 6 d or more" and IGF-1 "by 1.5- to 3-fold for 9–11 d"; estimated half-life 5.8–8.1 d; IGF-1 remained above baseline for up to 28 d after multiple doses. No p-values or CIs reported (JCEM). Combination: no numeric effect reported

Regulatory position: Neither agent is an approved medicine. FDA on ipamorelin: "FDA has not identified data to support the effectiveness of ipamorelin (free base) or ipamorelin acetate for the diagnosis or treatment of GHD"; "there are no data to support the effectiveness … for the proposed SC route of administration" (FDA PCAC). WADA 2026 List S2.2.4, naming "CJC-1295" among GHRH analogues and "ipamorelin" among growth hormone secretagogues (WADA). EU-specific legal determination: n.a.

Safety signal: Death reported in a CJC-1295 trial and FDA safety concerns for ipamorelin. A 2006 12-week phase II trial of CJC-1295 DAC in 192 HIV patients "reported 1 death from a myocardial infarction" approximately two hours after the 11th weekly dose; the trial physician judged it unrelated to treatment (FDA presentation). For ipamorelin, FDA concluded that "the use of ipamorelin-related bulk drug substances in compounding may raise safety concerns", listing adverse events of "hypokalemia, insomnia, hyperglycemia, nausea, vomiting, abdominal distention and death. However, it is unclear whether the two deaths were related to ipamorelin"; FDA further notes possible "behavioral reinforcing properties, which can contribute to development of addiction", possible negative effects on "reproductive health and pregnancy outcomes", and that nonclinical toxicity studies "were too limited in scope and duration" (FDA PCAC)

Sources: Teichman et al. 2006, *JCEM* Beck et al. 2014 FDA presentation FDA PCAC WADA

Follistatin-344 (delivered as rAAV1.CMV.huFS344 gene therapy)

Delivered as gene therapy, it adds muscle in people.

Claim as graded “Increases muscle mass in humans”

Evidence tier E · provenance grade P0 · direction Insufficient · retail band 6, Too little evidence yet. Derived by rule 3 - direction Insufficient. Group: SARMs and research peptides.

Best available evidence: Uncontrolled, open-label phase 1/2a gene-therapy trials in muscle disease — not muscle mass in healthy people, and with a published methodological rebuttal. (Mendell et al., BMD phase 1/2a; critique of the sIBM report)

Sample and design: n=6 Becker muscular dystrophy patients (Cohort 1 n=3 at 3×10¹¹ vg/kg/leg administered; Cohort 2 including patients 05 and 06 at 6×10¹¹ vg/kg/leg) (PubMed 25322757). The linked trial NCT01519349 enrolled 15 patients (9 sIBM, 6 BMD), "phase 1A", "open-label, single group assignment", with "no comparator or 'control' group" (critique, Molecular Therapy). A separate DMD trial, NCT02354781 (6 subjects, 2.4E12 vg/kg total), has no results posted (NCT02354781)

Effect as reported: Primary outcome was the 6-minute walk test, not muscle mass: patient 01 +58 m, patient 02 +125 m, patient 03 no change, patient 05 +108 m, patient 06 +29 m, patient 04 no improvement. No 95% CIs and no p-values reported. Histology showed "reduced endomysial fibrosis, reduced central nucleation, more normal fiber size distribution with muscle hypertrophy, especially at high dose" (PubMed 25322757). No numeric muscle-mass effect reported in humans

Regulatory position: Not an authorised medicine. WADA 2026 List S4.3 Agents Preventing Activin Receptor IIB Activation, which names "follistatin" among myostatin-binding proteins, prohibited at all times (WADA). EU legal determination: n.a. (searched EMA; nothing fetched)

Safety signal: No adverse effects were encountered in the BMD trial ("No adverse effects were encountered") (PubMed 25322757). Evidence-integrity signal rather than a harm signal: a published critique of the sIBM report documents that the analysis used "a post hoc-defined primary outcome measure", was "an unstated interim analysis at a post hoc chosen time-point", compared against an unmatched 8-patient clinic cohort, and confounded the gene therapy with "high-dose prednisone for approximately 60 days", a monitored exercise programme, and placebo effects — "at least 4 potentially therapeutic interventions". One subject had "no detectable FS344 DNA present" in post-treatment biopsies (PMC5628928)

Sources: Mendell et al., BMD phase 1/2a critique of the sIBM report NCT02354781 WADA

IGF-1 LR3 (Long R3 IGF-1)

It grows muscle in people.

Claim as graded “Increases muscle growth in humans”

Evidence tier E · provenance grade P0 · direction Insufficient · retail band 6, Too little evidence yet. Derived by rule 3 - direction Insufficient. Group: SARMs and research peptides.

Best available evidence: None. No human trial of IGF-1 LR3 for muscle growth was located. The only IGF-1-related human data fetched concern recombinant human IGF-1/IGFBP-3 in an anti-doping detection study, not IGF-1 LR3 and not a muscle-growth endpoint (Nicholls et al. review, PMC7913862)

Sample and design: n.a. — zero human participants for this claim. Searched ClinicalTrials.gov via web search for "IGF-1 LR3" human trials; the only registered IGF-1 trial surfaced was of native IGF-1 in autism (NCT01970345), not IGF-1 LR3 and not muscle growth

Effect as reported: no numeric effect reported

Regulatory position: Not an authorised medicine. WADA 2026 List S2.3, naming "Insulin-like growth factor 1 (IGF-1, mecasermin) and its analogues" — IGF-1 LR3 is an analogue — prohibited at all times (WADA). EU legal determination: n.a.

Safety signal: Elevated IGF-1 is associated with increased cancer risk in observational human data. In 206,263 UK Biobank women, those in the top 20% of IGF-1 concentration had a "1.24-fold increased chance of developing breast cancer" versus the bottom 20%, and Mendelian randomisation across 122,977 cases and 105,974 controls found risk "increased by 1.05 for every additional genetically predicted 5 nmol/L of IGF-1" (PMC7913862). This is an association for circulating IGF-1, not a demonstrated harm of administered IGF-1 LR3

Sources: Nicholls et al. review, PMC7913862 NCT01970345 WADA

Vitamin C (ascorbic acid), high-dose supplementation [L-ascorbic acid]

Taken regularly, it shortens how many days a cold lasts in adults, counted from the symptom diaries they kept themselves.

Claim as graded “"Shortens your cold" — reduces the mean duration of naturally occurring common cold episodes in adults taking ≥0.2 g/day regularly, measured as days of illness from participant-kept symptom records”

Evidence tier A · provenance grade P1 · direction Contested · retail band 4, Experts disagree. Derived by rule 2 - direction Contested. Group: Vitamins, minerals and recovery.

Best available evidence: Cochrane systematic review and meta-analysis of placebo-controlled trials (Hemilä & Chalker, Cochrane Database Syst Rev, PMC8078152)

Sample and design: Regular supplementation: 29 comparisons, 11,306 participants for incidence; 31 comparisons, 9,745 cold episodes for duration, of which 17 adult entries representing 21 trial arms, 7,215 illness episodes and 14 child comparisons, 2,530 episodes. Therapeutic (dosing started at symptom onset): 7 entries incorporating 10 trial arms, 3,249 cold episodes

Effect as reported: Duration, regular supplementation, adults: 7.7% shorter (95% CI 3.7% to 12%), i.e. MD −7.72 (95% CI −11.76 to −3.69); children 14.2% (7.3% to 21%); restricted to randomised double-blind trials, adults 7% (3% to 11%). Incidence in the general community: RR 0.97 (95% CI 0.94 to 1.00), 24 entries, 10,708 participants and at ≥1 g/day RR 0.98 (95% CI 0.95 to 1.01); under short-term severe physical stress RR 0.48 (95% CI 0.35 to 0.64), 5 studies, 598 participants. Therapeutic dosing at onset is null: duration MD −2.90 (95% CI −8.20 to 2.39); severity SMD −0.07 (95% CI −0.15 to 0.01). No p-values stated on the fetched page

Regulatory position: EU authorised Art.13(1) claims exist but are narrower than what is sold. Authorised: "Vitamin C contributes to the normal function of the immune system" — condition ["The claim may be used only for food which is at least a source of vitamin C as referred to in the claim SOURCE OF [NAME OF VITAMIN/S] AND/OR [NAME OF MINERAL/S] as listed in the Annex to Regulation (EC) No 1924/2006."](https://ec.europa.eu/food/food-feed-portal/backend/claims/files/euregister.pdf) (EFSA 2009;7(9):1226, 2010;8(10):1815; Reg. (EU) 432/2012). Also authorised: "Vitamin C contributes to maintain the normal function of the immune system during and after intense physical exercise" — condition "The claim may be used only for food which provides a daily intake of 200 mg vitamin C…". No authorised claim mentions the common cold, cold duration or cold severity; the nearest respiratory wording, ascorbic acid "Soothing for mouth and throat / Reliefs in case of tickle in the throat and pharynx…", is Non-authorised because "this claimed effect for this food has not been substantiated"

Safety signal: Tolerable Upper Intake Level for adults 2,000 mg/day; "Long-term intakes of vitamin C above the UL may increase the risk of adverse health effects". "High doses of vitamin C can lead to diarrhea, nausea, abdominal cramps, and other gastrointestinal disturbances". "The best evidence that vitamin C contributes to kidney stone formation is in patients with pre-existing hyperoxaluria". In hereditary haemochromatosis, "chronic consumption of high doses of vitamin C could exacerbate iron overload and result in tissue damage"

Sources: PMC8078152 EU Register PDF ODS vitamin C, health professional ODS vitamin C, consumer

N-acetylcysteine [NAC, acetylcysteine]

It cuts how often adults with a long-term lung disease get a flare-up.

Claim as graded “"Protects your airways" — reduces the incidence of exacerbations in adults with COPD or with chronic bronchitis / pre-COPD, measured as the total number of exacerbations”

Evidence tier B · provenance grade P2 · direction Supported · retail band 2, Moderate evidence. Derived by rule 7 - tier A/B with provenance P2. Group: Vitamins, minerals and recovery.

Best available evidence: Systematic review and meta-analysis of RCTs, PROSPERO-registered (Cazzola 2024, PubMed 38555190); second meta-analysis of AECOPD trials (PubMed 34237968)

Sample and design: Twenty studies were included, of which seven evaluated NAC in patients with symptoms of CB/pre-COPD as entry criterion; total participants not stated on the fetched page. Second source: a total of 15 included literatures; total participants not stated

Effect as reported: Exacerbation incidence vs placebo, COPD: IRR = 0.76 (95% CI 0.59–0.99); chronic bronchitis/pre-COPD: IRR = 0.81 (95% CI 0.69–0.95); improvement in symptoms and/or quality of life: OR = 3.47 (95% CI 1.92–6.26). No p-values and no I² stated on the fetched page. Second meta-analysis, AECOPD: symptom improvement rate risk rate 1.09 (1.04–1.14), P<0.0001, I²=0%; FEV1 MD 30.63 (25.48–35.78), P<0.0001, I²=92%; FEV1/FVC MD 30.42 (24.00–36.85), P<0.0001, I²=93%

Regulatory position: The marketed supplement claim is refused in the EU. N-acetyl-l-cysteïne — "Adequate supply contributes to glutathione homeostasis /restoring glutathione levels with cysteine helps to keep the redox state in balance /building glutathione levels with cysteine support the body's natural defense system /building glutathione levels with cysteine supports the detoxification function of your liver" — Non-authorised, reason "this claimed effect for this food has not been substantiated" (EFSA 2010;8(10):1795, entry 1745). A product-specific detox claim combining "sodium alginate, n-acetyl cysteine and piperine" is likewise Non-authorised (EFSA 2011;9(6):2248). No authorised EU health claim for NAC was located

Safety signal: Hepatotoxicity signal is negative: LiverTox likelihood score E — unlikely cause of clinically apparent liver injury; "Since approval of the oral and intravenous forms of acetylcysteine, there have been no published reports of hepatotoxicity". Non-hepatic signals are real and quantified: in 1,329 patients given oral NAC, diarrhea was common (44%) as was nausea and vomiting (35%); in 60 patients given intravenous NAC, 23% had an anaphylactoid reaction (flushing, rash, angioedema, wheezing); in 179 patients, 32 (18%) had side effects that were mostly transient flushing and rash; in 57 children with non-acetaminophen acute liver failure, adverse events (11%) included rash, arrhythmias, edema and bronchospasm

Sources: PubMed 38555190 PubMed 34237968 EU Register PDF LiverTox Acetylcysteine

Glycine [aminoacetic acid]

Taken before bed, it gives adults who sleep badly better-rated sleep and less morning grogginess.

Claim as graded “"Improves your sleep quality" — improves subjective sleep quality and next-morning fatigue in adults with self-reported poor sleep (PSQI ≥6) taking glycine before bedtime”

Evidence tier D · provenance grade P1 · direction Insufficient · retail band 6, Too little evidence yet. Derived by rule 3 - direction Insufficient. Group: Vitamins, minerals and recovery.

Best available evidence: Scoping/systematic review of 47 articles in which meta-analysis was not possible (Aguayo-Cerón / Front Nutr-type review, PMC10828290); underlying human sleep trials described in a pharmacology review (Bannai & Kawai, J Pharmacol Sci, J-STAGE); one crossover trial in sleep-restricted workers (Bannai 2012, PMC3328957)

Sample and design: Review: 47 articles describing 50 studies were included; most studies (42/50) were randomised controlled trials; aggregate participant total not reported. Sleep-specific human evidence is three small trials: 19 female volunteers, randomized double-blinded crossover; 11 volunteers (8 female, 3 male), randomized single-blinded crossover with polysomnography; 12 volunteers, open trial. Fourth trial: 10 healthy male volunteers, 7 analysed, randomized single-blinded crossover

Effect as reported: `no numeric effect reported` — a meta-analysis combining the extracted data to ascertain the overall effect of glycine administration on the characteristics for each physiological system could not be performed due to the large heterogeneity and nature of reported outcomes and statistical presentation of the data, and formal statistical analysis was not conducted and results are interpreted on reported p-values. Narrative findings only: 3 g of glycine before bedtime significantly improved the feeling of fatigue the next morning; polysomnography revealed a stabilized sleep state and a shortened latency to both the sleep onset and slow-wave sleep, with no alterations in the sleep architecture — no numerical values for sleep latency or slow-wave sleep, and no p-values, confidence intervals, effect sizes, means or standard deviations, are stated

Regulatory position: No entry for "glycine" is found in the EU Register of nutrition and health claims — no authorised health claim, and the substance has not been evaluated there for a sleep claim

Safety signal: No serious signal found in fetched sources. Glycine (9 g) administered during the day did not induce sleepiness and had no adverse effects; no serious side effects have been observed with the administration of 31 g/day of glycine

Sources: PMC10828290 J-STAGE Bannai & Kawai PMC3328957 PMC11510825 EU Register PDF

L-theanine [γ-glutamylethylamide, green-tea amino acid]

A single dose takes the edge off short-term stress in healthy adults.

Claim as graded “"Takes the edge off" — reduces acute stress in healthy adults after a single oral dose”

Evidence tier B · provenance grade P1 · direction Contested · retail band 4, Experts disagree. Derived by rule 2 - direction Contested. Group: Vitamins, minerals and recovery.

Best available evidence: Systematic review and meta-analysis of RCTs (Cognitive and affective effects of L-Theanine, PubMed 42410082); supporting narrative systematic review (Williams 2020, PubMed 31758301); PROSPERO-registered cognition-only meta-analysis (PMC12609247)

Sample and design: 31 randomised controlled trials, n = 1168, in healthy and clinical populations, all comparing oral L-theanine with placebo. Narrative review: 9 peer-reviewed journal articles; participants not stated. Cognition meta-analysis: five RCTs, 148 healthy adults; all were randomized, placebo-controlled, double-blind studies

Effect as reported: Primary outcome, acute stress after a single dose in healthy adults: SMD = 0.31 — the fetched page states no 95% CI and no p-value for this estimate, and reports the effect was "modest (SMD = 0.31) and largely influenced by studies with a high risk of bias". Secondary: no significant effect on fatigue was observed; anxiety effects were inconsistent and non-significant except for one study on psychotic anxiety (SMD = 0.54); depressive symptoms after a single dose, excluding one outlier, SMD = 0.69 (95% CI 0.13–1.25); choice reaction time SMD = 0.51 (95% CI 0.25–0.77)

Regulatory position: Explicitly non-authorised in the EU, including for the exact wording sold. L-Theanine — "-help relaxation without drowsiness. -Stress relief. -Physical stress relief. -Relax from fatigue. -Support relation for optimal mental and physical well-being." — Non-authorised (health relationship: alleviation of psychological stress), reason "this claimed effect for this food has not been substantiated", EFSA 2011;9(6):2238, entry 1598. Also non-authorised: "helps to maintain an optimal relaxation; helps to support the relaxation ; helps to maintain a healthy sleep" (entry 1737) and "- Help learning performance. - Help to improve concentration. - Help to improve attention. - Sports support" (entry 1600)

Safety signal: No serious signal found. No serious adverse events were reported; safety was assessed by comparing dropout rates and reasons between L-theanine and placebo groups, with numbers not stated on the page

Sources: PubMed 42410082 PubMed 31758301 PMC12609247 EU Register PDF

Valerian (Valeriana officinalis) [valerian root, Valerianae radix]

It gives adults with insomnia or poor sleep better-rated sleep.

Claim as graded “"Helps you sleep better" — improves subjective sleep quality in adults with insomnia or self-reported poor sleep”

Evidence tier C · provenance grade P1 · direction Contested · retail band 4, Experts disagree. Derived by rule 2 - direction Contested. Group: Vitamins, minerals and recovery.

Best available evidence: Two systematic reviews with meta-analysis of randomised placebo-controlled trials (Bent 2006, Am J Med, PMC4394901; Fernández-San-Martín 2010, Sleep Med, PubMed 20347389)

Sample and design: Sixteen randomized, controlled trials, 1093 patients; the pooled dichotomous analysis rested on six of them. Second review: eighteen RCTs were selected, of which eight had a score of 5 on Jadad's scale; total participants not stated

Effect as reported: Dichotomous "sleep improved or not": relative risk of improved sleep = 1.8 (95% CI 1.2–2.9), no p-value reported; after excluding one study relative risk = 1.9 (95% CI 1.6–2.3), P value for heterogeneity = .3; publication bias by Kendall's tau (P < .01). Second review, same dichotomous construct: relative risk of SQ of 1.37 (95% CI 1.05–1.78) — but the quantitative outcomes are null: sleep-onset latency mean difference 0.70 min (95% CI −3.44 to 4.83) and VAS sleep quality SMD −0.02 (95% CI −0.35 to 0.31); no p-values stated

Regulatory position: Licensed herbal medicine, not a food claim. EMA HMPC: "The efficacy of a specific preparation of Valerian root in treating mild nervous tension and sleep disorders is based on its 'well-established use'", meaning "documented evidence of their medicinal use covering at least 10 years in the EU with published clinical studies to support the efficacy and safety when used for the relief of mild nervous tension and to improve the quality of sleep and the time to fall asleep"; other preparations (juice, water extracts, methanolic extracts, tinctures) rest on "traditional use", i.e. "although there is insufficient evidence from clinical trials, the effectiveness of these herbal medicines is plausible". No entry for "valerian" or "Valeriana" is found in the EU Register of nutrition and health claims

Safety signal: Yes — hepatotoxicity. LiverTox likelihood score C: probable rare cause of clinically apparent liver injury. "Valerian has been implicated in a small number of cases of clinically apparent liver injury, but usually in combination with other botanicals such as skullcap or black cohosh"; "Severe cases with features of hepatic failure have been described". Published case: a 57-year-old woman developed jaundice 3 weeks after starting a valerian-containing herbal, with serum bilirubin 18.4 mg/dL and ALT 1165 U/L, INR=1.9, who developed ascites and hepatic encephalopathy and had only incomplete recovery at 10 months. Non-hepatic: "sedation, dizziness and withdrawal symptoms on stopping"

Sources: PMC4394901 PubMed 20347389 EMA Valerianae radix EU Register PDF LiverTox Valerian

Glucosamine with chondroitin [glucosamine sulphate/hydrochloride + chondroitin sulphate]

It eases pain in adults with arthritis of the hip or knee.

Claim as graded “"Relieves joint pain" — reduces pain intensity in adults with osteoarthritis of the hip or knee”

Evidence tier B · provenance grade P1 · direction Not supported · retail band 5, Evidence says no. Derived by rule 1 - direction Not supported. Group: Vitamins, minerals and recovery.

Best available evidence: Network meta-analysis of large-scale randomised trials (Wandel 2010, BMJ, PubMed 20847017)

Sample and design: 10 trials, 3803 patients

Effect as reported: Pain intensity vs placebo: glucosamine plus chondroitin −0.5 cm (95% credible interval −0.9 to 0.0 cm) on a 10 cm visual analogue scale, against a prespecified minimal clinically important difference of −0.9 cm; glucosamine alone −0.4 cm (95% CrI −0.7 to −0.1); chondroitin alone −0.3 cm (95% CrI −0.7 to 0.0). Joint space: the differences in changes in minimal width of joint space were all minute, with 95% credible intervals overlapping zero. Funding effect: industry independent trials showed smaller effects than commercially funded trials (P=0.02 for interaction). No I² reported on the fetched page

Regulatory position: Refused in the EU. Art.13(5): "Glucosamine contributes to the maintenance of normal joint cartilage" (Merck Consumer Healthcare, Question No EFSA-Q-2011-01113) and "Glucosamine contributes to the protection of joint cartilage exposed to excessive motion or loading and helps to improve the range of motion in joints" (Béres Pharmaceuticals, EFSA-Q-2011-00907) are both listed under "Rejected health claims" in Commission Regulation (EU) No 1066/2013, because "a cause and effect relationship had not been established between the consumption of glucosamine and the claimed effect". EFSA's own conclusion: "a cause and effect relationship has not been established between the consumption of glucosamine and maintenance of normal joint cartilage in individuals without osteoarthritis". Art.13(1) chondroitin joint wordings are also Non-authorised, e.g. "Renowned for helping maintain joint mobility and flexibility. Chondroitin (and glucosamine) may help to support healthy knees…" (EFSA 2009;7(9):1262)

Safety signal: No serious signal found in fetched sources — the fetched network meta-analysis reports no adverse-event data

Sources: PubMed 20847017 Reg. (EU) 1066/2013 EFSA Journal 2691 EU Register PDF

Boswellia serrata (frankincense extract) [Indian frankincense, salai guggul, boswellic acids]

It eases pain and stiffness in adults with knee arthritis.

Claim as graded “"Eases stiff, painful knees" — reduces knee pain and stiffness in adults with knee osteoarthritis”

Evidence tier C · provenance grade P1 · direction Supported · retail band 3, Widely studied, poorly measured. Derived by rule 5 - provenance cap at P1 or below. Group: Vitamins, minerals and recovery.

Best available evidence: Systematic review and meta-analysis of RCTs, conducted to a registered protocol (Yu 2020, BMC Complement Med Ther, PMC7368679)

Sample and design: Seven trials involving 545 patients; 7 RCTs with 545 participants, conducted in India, Armenia, Iran and Italy; protocol CRD42018086785

Effect as reported: VAS at end of treatment: WMD −8.33 (95% CI −11.19 to −5.46), P<0.00001, I² = 94%; WOMAC pain WMD −14.22 (95% CI −22.34 to −6.09), P = 0.0006, I² = 99%; WOMAC stiffness WMD −10.04 (95% CI −15.86 to −4.22), P = 0.0007, I² = 97%; WOMAC function WMD −10.75 (95% CI −15.06 to −6.43), P<0.00001, I² = 93%; Lequesne Index WMD −2.27 (95% CI −3.08 to −1.45), P<0.00001, I² = 47%. Adverse events RR 0.63 (95% CI 0.22–1.83), P = 0.39

Regulatory position: No entry for Boswellia serrata is found in the EU Register of nutrition and health claims — no authorised health claim; not evaluated in that register. No EMA herbal monograph was fetched

Safety signal: No hepatic signal: LiverTox likelihood score E — unlikely cause of clinically apparent liver injury; Boswellia serrata extract has not been linked to serum enzyme elevations during therapy. Gastrointestinal signal only: in 42 patients with ulcerative colitis, 18% on Boswellia had gastrointestinal side effects such as heartburn, anorexia, nausea or abdominal pain; overall "side effects are few and largely mild and transient gastrointestinal symptoms of nausea, diarrhea or constipation"

Sources: PMC7368679 EU Register PDF LiverTox Boswellia Serrata

Iron supplementation in iron-deficient athletes [ferrous sulfate/fumarate/gluconate; oral and intravenous iron]

It restores aerobic fitness in endurance athletes who are low on iron but not anaemic.

Claim as graded “"Restores your endurance" — increases maximal aerobic capacity (VO₂max) in iron-deficient non-anaemic endurance athletes”

Evidence tier C · provenance grade P2 · direction Contested · retail band 4, Experts disagree. Derived by rule 2 - direction Contested. Group: Vitamins, minerals and recovery.

Best available evidence: Systematic review and meta-analysis (Burden 2015, Br J Sports Med, PubMed 25361786); contrasting systematic review without pooling (Rubeor 2018, Sports Health, PMC6116100)

Sample and design: Meta-analysis: seventeen eligible studies were identified from online databases; total participants not stated on the fetched page. Second review: 12 studies with a total of 283 participants

Effect as reported: VO₂max: Hedges' g = 0.610 (95% CI 0.399 to 0.821), p<0.001. Iron-status co-outcomes: serum ferritin g = 1.088 (0.914–1.263), p<0.001; serum iron g = 1.004 (0.828–1.181), p<0.001; transferrin saturation g = 0.741 (0.564–0.919), p<0.001; haemoglobin g = 0.695 (0.533–0.836), p<0.001. No I² reported. The second review reaches the opposite conclusion on performance: supplementing IDNA athletes with iron improved performance in 6 studies (146 participants) and did not improve performance in the other 6 studies (137 participants), and the results of our review are equivocal and do not support the conclusion that iron supplementation consistently improves performance in endurance IDNA athletes

Regulatory position: EU authorised Art.13(1) claims exist but are narrower than the performance claim sold: "Iron contributes to normal oxygen transport in the body", "Iron contributes to normal formation of red blood cells and haemoglobin" and "Iron contributes to the reduction of tiredness and fatigue" — each conditioned on ["The claim may be used only for food which is at least a source of iron as referred to in the claim SOURCE OF [NAME OF VITAMIN/S] AND/OR [NAME OF MINERAL/S] as listed in the Annex to Regulation (EC) No 1924/2006."](https://ec.europa.eu/food/food-feed-portal/backend/claims/files/euregister.pdf) (EFSA 2009;7(9):1215, 2010;8(10):1740; Reg. (EU) 432/2012). No authorised claim mentions aerobic capacity, endurance or athletic performance

Safety signal: Yes — overdose and corrosive toxicity. "High doses of supplemental iron (45 mg/day or more) may cause gastrointestinal side effects"; "Acute intakes of more than 20 mg/kg iron (about 1,365 mg iron for a person weighing 150 pounds) from supplements or medicines can lead to corrosive necrosis of the intestine, which might lead to fluid and blood loss, shock, tissue damage, and organ failure"; "In severe cases (e.g., one-time ingestions of 60 mg/kg…), overdoses of iron can lead to multisystem organ failure, coma, convulsions, and even death"; "Between 1983 and 2000, at least 43 U.S. children died from ingesting supplements containing high doses of iron (36–443 mg iron/kg body weight)". Also "Supplements containing 25 mg iron or more can reduce zinc absorption and plasma zinc concentrations", and iron supplementation in anaemia of chronic disease "might increase the risk of infection and cardiovascular events and could cause tissue damage"

Sources: PubMed 25361786 PMC6116100 EU Register PDF ODS Iron, health professional

Probiotics (Lactobacillus / Bifidobacterium strains) [live micro-organisms, single- and multi-strain]

They cut how many people catch at least one cold or throat infection — in children, adults and older people.

Claim as graded “"Keeps you from getting sick this winter" — reduces the number of people who develop at least one acute upper respiratory tract infection, in children, adults and older people”

Evidence tier B · provenance grade P2 · direction Supported · retail band 2, Moderate evidence. Derived by rule 7 - tier A/B with provenance P2. Group: Vitamins, minerals and recovery.

Best available evidence: Cochrane systematic review and meta-analysis (Zhao 2022, Cochrane Database Syst Rev CD006895.pub4; PubMed 36001877)

Sample and design: 23 individual RCTs and one cluster-RCT, a total of 6950 participants, including children (aged from one month to 11 years old), adults (mean age 37.3), and older people (mean age 84.6 years); data from 23 trials could be meta-analysed

Effect as reported: At least one URTI event: RR 0.76, 95% CI 0.67 to 0.87; P < 0.001; 16 studies, 4798 participants; low-certainty evidence. At least three events: RR 0.59, 95% CI 0.38 to 0.91; P = 0.02; 4 studies, 763 participants; moderate-certainty evidence. Incidence rate: rate ratio 0.82, 95% CI 0.73 to 0.92, P = 0.001; 12 studies, 4364 participants; low-certainty evidence. Mean episode duration: MD −1.22 days, 95% CI −2.12 to −0.33; P = 0.007; 6 studies, 2406 participants; low-certainty evidence. Antibiotic prescriptions: RR 0.58, 95% CI 0.42 to 0.81; P = 0.001; 6 studies, 1548 participants. I² not stated on the fetched pages

Regulatory position: Explicitly refused in the EU for exactly this claim. Art.13(5): "Daily consumption of live Lactobacillus casei strain Shirota as present in a fermented milk product helps maintain the upper respiratory tract defences by helping to support immune functions" — Non-authorised (Q-2010-00137, Commission Regulation (EU) No 1171/2011), reason "this claimed effect for this food has not been substantiated". Also refused: "Lactobacillus GG helps to maintain defence against intestinal pathogens" (Valio Ltd, EFSA-Q-2010-01028, listed under "Rejected health claims" in Reg. (EU) No 379/2012) and LACTORAL immunity claims (Reg. (EC) No 1024/2009). No authorised Art.14 or Art.13(5) entry for Lactobacillus or Bifidobacterium relating to respiratory infections or immune defence was found

Safety signal: Yes — one confirmed fatal contamination case in a preterm infant. CDC MMWR: a hospital reported "a fatal case of gastrointestinal mucormycosis in a preterm infant" who "had received a dietary supplement, ABC Dophilus Powder, for 7 days, beginning on day 1 of life"; "Unopened bottles of ABC Dophilus Powder from the lot received by the infant were cultured… the samples yielded Rhizopus species", confirmed by CDC as "Rhizopus oryzae, the same species of fungus recovered from the unopened dietary supplement"; mortality for this infection is stated as 85%, and "no additional cases of gastrointestinal mucormycosis in neonates have been identified to date". In the trial literature, adverse events were not increased: RR 1.02, 95% CI 0.90 to 1.15; P = 0.79; 8 studies, 2456 participants, and adverse events from probiotics were minor, and most commonly gastrointestinal symptoms

Sources: Cochrane CD006895.pub4 PubMed 36001877 EU Register PDF Reg. (EU) 379/2012 CDC MMWR 64(06)

Saffron (Crocus sativus) extract [saffron stigma extract, crocin/safranal-standardised]

It lowers depressive symptoms in adults with depression or clear depressive symptoms.

Claim as graded “"Lifts low mood" — reduces depressive symptoms in adults with depression or clinical depressive symptoms”

Evidence tier B · provenance grade P2 · direction Contested · retail band 4, Experts disagree. Derived by rule 2 - direction Contested. Group: Vitamins, minerals and recovery.

Best available evidence: Systematic review and meta-analysis of RCTs (Marx 2019, Nutrition Reviews 77(8):557); second meta-analysis (Toth 2019 / PMC6503633)

Sample and design: Twenty-three studies were included; a total of 1237 participants were enrolled, with 30–128 participants in each study; trials ran 4–12 weeks; most studies were conducted in Iran (n = 21/23) and thirteen studies were conducted by the same research group. Second meta-analysis: eight studies, all described in their titles as double-blind

Effect as reported: Depressive symptoms, saffron vs placebo: Hedges' g = 0.99, 95% CI 0.61–1.37, P < 0.001; n = 14 studies, n = 716 participants; Q = 71.8, I² = 81.9%; trim-and-fill corrected g = 1.14 (0.74–1.52); excluding one outlier g = 0.84 (0.53–1.16), P < 0.001, I² = 73.8%. Saffron vs antidepressant medication: Hedges' g = −0.17, 95% CI 0.50–0.17, P = 0.33; n = 5 studies, n = 210; I² = 35.1%. Publication bias is explicit: "Egger's regression test found strong evidence of publication bias (Intercept = 6.99, P = 0.007)", still present after removing the outlier (intercept = 4.957, P = 0.043). Second meta-analysis, saffron vs placebo: SMD −0.86 (95% CI −1.73 to 0.00), heterogeneity 87%; vs fluoxetine SMD 0.11 (95% CI −0.20 to 0.43)

Regulatory position: No entry for "saffron" or "Crocus sativus" is found in the EU Register of nutrition and health claims — no authorised health claim; not evaluated in that register. No EMA herbal monograph was fetched

Safety signal: No serious signal found in fetched sources. The fetched meta-analysis does not report numerical adverse-event data; no LiverTox entry for saffron was located this session

Sources: Nutrition Reviews 77(8):557 PMC6503633 EU Register PDF