MK-677 (Ibutamoren): An Evidence Ledger

A spiroindoline, not a peptide

MK-677 is a small synthetic molecule built on a spiro[indoline-3,4'-piperidine] core carrying a methanesulfonyl group on the indoline nitrogen. PubChem records the free base as CID 178024, C27H36N4O5S, 528.7 g/mol, CAS 159634-47-6; the mesylate salt, which is what PubChem returns for the name MK-677, is CID 6450830, C28H40N4O8S2, 624.8 g/mol, CAS 159752-10-0. There is no amino acid sequence because there is no peptide. Merck's medicinal chemistry group described the compound in 1995 under the internal number L-163,191, reporting release of growth hormone from cultured rat pituitary cells with an EC50 of 1.3 plus or minus 0.09 nM.

Selectivity was part of the original case for taking it forward. Patchett and colleagues reported growth hormone elevation in dogs after oral doses as low as 0.125 mg/kg, with no significant effect on plasma aldosterone, luteinising hormone, thyroxine or prolactin at 1 mg/kg, and only modest increases in cortisol. Holst and colleagues later characterised the receptor interaction directly: MK-677 bound the ghrelin receptor with an affinity of 6.5 nM and activated calcium mobilisation, inositol phosphate turnover, CRE- and SRE-controlled transcription and arrestin mobilisation at potencies between 0.2 and 1.4 nM, behaving as a simple agonist and a neutral modulator of ghrelin signalling where the benzolactam L-692,429 was positive and GHRP-6 negative.

Two labels travel with this compound and neither survives contact with the chemistry. It is sold as a peptide, and it is filed alongside selective androgen receptor modulators. Its target is the growth hormone secretagogue receptor GHS-R1a — the receptor for ghrelin — several steps upstream of anything androgenic. A PubMed search pairing ibutamoren with androgen receptor on 18 August 2026 returned six records — analytical chemistry, market-surveillance surveys, two case reports, one review and one in vitro receptor-assay panel — none of which reports androgen receptor binding or transactivation by the compound at compound level.

What a dose did to the axis

Chapman and colleagues gave 32 healthy adults aged 64 to 81 placebo or 2, 10 or 25 mg orally once daily across two study periods of 14 and 28 days, sampling blood every 20 minutes for 24 hours. At 25 mg per day, mean 24-hour growth hormone concentration rose 97 plus or minus 23 per cent against baseline. Serum IGF-I moved from 141 plus or minus 21 micrograms per litre at baseline to 219 plus or minus 21 at two weeks and 265 plus or minus 29 at four weeks. Prolactin rose 23 per cent and stayed inside the stated normal range.

How the pulse pattern changed is where two studies diverge. Chapman reported increased pulse height and increased interpulse nadir concentrations with no significant change in the number of pulses. Copinschi and colleagues, dosing nine healthy young men at bedtime for seven days in a three-period crossover, found that the total amount of growth hormone secreted was similar under placebo, 5 mg and 25 mg, and that pulse frequency increased — mainly through additional low-amplitude pulses. Age, dosing time, duration and assay all differ between the two experiments. This page records the disagreement rather than averaging it.

Deficient populations were tested separately. Chapman's 1997 study of nine severely growth-hormone-deficient men aged 17 to 34 found IGF-I rose 52 plus or minus 20 per cent on 10 mg daily and 79 plus or minus 9 per cent on 50 mg daily over four days. The 24-hour mean growth hormone response — not the IGF-I response — was larger in the subjects least deficient at baseline; the paper states that IGF-I responses were not significantly related to any baseline measurement. Codner and colleagues gave 18 prepubertal children with idiopathic deficiency 0.2 or 0.8 mg/kg per day for seven to eight days and recorded a median IGF-I rise of 12 micrograms per litre (range minus 4 to 116) at the higher dose, with no change in prolactin, glucose, thyroid hormones, peak cortisol or 24-hour urinary free cortisol.

Glucose is the finding that repeats

Almost every study that looked found something. Svensson and colleagues randomised 24 obese men aged 18 to 50 to 25 mg daily or placebo for eight weeks. Fasting glucose and insulin were unchanged, but an oral glucose tolerance test showed impaired glucose homeostasis at both two and eight weeks. Fat-free mass increased on dual-energy x-ray absorptiometry and on a four-compartment model, while total and visceral fat did not change significantly. Basal metabolic rate was raised at two weeks and not at eight. Chapman's older-adult study is blunter still: fasting glucose moved from 5.4 plus or minus 0.3 to 6.8 plus or minus 0.4 mmol/L at four weeks.

Nass and colleagues, running the longest exposure on record, recorded a mean fasting glucose increase of 0.3 mmol/L over 12 months and a decrease in insulin sensitivity. That is a small number, and it sat alongside a 2.7 kg gain in body weight and an increase in appetite that the investigators described as subsiding within a few months. The direction is consistent across the 1995-2022 record; the magnitude is not, and it appears to depend heavily on how long the exposure ran and how glucose was challenged.

One recent trial looked at the liver. A single-arm open-label study at Massachusetts General Hospital gave 25 mg daily for six months to 12 adults with a body mass index of 25 or above and a radiological or histological diagnosis of fatty liver disease, then posted its results to ClinicalTrials.gov. Intrahepatic lipid rose by 3.9 percentage points (SD 6.7). LiverMultiScan corrected T1 rose 39 ms (SD 44) against a stated normal range of 633 to 794 ms. Alanine aminotransferase rose 8 U/L (SD 18). Hyperglycaemia was recorded in 4 of 12 participants and increased appetite in 7. Twelve participants, no control arm, historical comparators, and no publication indexed in PubMed.

The two-year trial, and what it did not show

Nass and colleagues ran a two-year double-blind randomised placebo-controlled modified-crossover trial in 65 healthy adults aged 60 to 81 at a university general clinical research centre, dosing 25 mg orally once daily. Fat-free mass was a primary endpoint at one year. It fell 0.5 kg (95% CI minus 1.1 to 0.2) on placebo and rose 1.1 kg (0.7 to 1.5) on MK-677, P less than 0.001. Body weight rose 2.7 kg against 0.8 kg. Body cell mass, read as intracellular water, moved 0.8 kg against minus 1.0 kg. Two-year exploratory analyses confirmed the one-year results.

Abdominal visceral fat was the other primary endpoint and did not differ between groups. Neither did total fat mass. Limb fat rose more on drug than on placebo, 1.1 kg against 0.24 kg. The results section also states, in one sentence: increased fat-free mass did not result in changes in strength or function. Isokinetic strength, function and quality of life were all measured endpoints, and the authors listed insufficient power for functional endpoints among the limitations.

Cortisol did move here, by 47 nmol/L (95% CI 28 to 71), P = 0.020 — a direction the shorter studies had not seen. LDL cholesterol fell 0.14 mmol/L from baseline with no between-group difference in total or HDL cholesterol. Bone mineral density changes were described as consistent with increased bone remodelling. Reported side effects were an increase in appetite, transient mild lower-extremity oedema and muscle pain. That is the profile of the most favourable trial in the file, run in healthy volunteers under research-centre supervision.

Where the programme stopped

Bone came first. Murphy and colleagues randomised 292 women aged 64 to 85 with low femoral neck density into four arms for 12 months, following density to 18 months. MK-677 raised osteocalcin by 22 per cent and urinary N-telopeptide by 41 per cent — both formation and resorption markers, in the same direction. Combined with alendronate it raised femoral neck density 4.2 per cent against 2.5 per cent for alendronate alone. No similar enhancement appeared at the lumbar spine, total hip or total body. Svensson's obese-male cohort had shown the same two-sided turnover picture at eight weeks.

Alzheimer's disease was the largest attempt. Sevigny and colleagues randomised 563 patients with mild to moderate disease to 25 mg or placebo for 12 months, of whom 416 completed. Serum IGF-I rose 60.1 per cent at six weeks and 72.9 per cent at 12 months, which the authors read as target engagement. There were no significant differences on the clinician's interview-based impression of change, the ADAS-Cog, the activities-of-daily-living scale or the clinical dementia rating sum of boxes. The registry record for the corresponding Merck protocol lists 512 participants as actual enrolment against the paper's 563 randomised, posts no results, and neither document explains the gap.

Hip fracture produced the safety signal. Bach and colleagues enrolled 161 patients aged 65 and over across 13 centres for six months of daily treatment: IGF-I rose 84 per cent against 17 per cent on placebo, and no significant difference emerged in functional performance measures or in the overall Sickness Impact Profile score. Adunsky and colleagues then ran a phase 2b trial in 123 elderly hip-fracture patients, 62 on 25 mg and 61 on placebo. Gait speed differed by 0.7 in the means (95% CI 0.17 to 1.28; P = 0.011). Stair-climbing power did not (P = 0.292), nor did most other functional measures. The trial was terminated early for a congestive heart failure signal in a limited number of patients, and the authors wrote that the compound has an unfavourable safety profile in that population.

The same molecule, a different name, a different market

Development did not end with Merck. ClinicalTrials.gov holds 14 studies naming MK-677, MK-0677, ibutamoren or LUM-201, and the registry states in plain text that LUM-201 is ibutamoren mesylate. Four of the 14 have posted results. In one phase 2 trial, 104 children with growth hormone deficiency received LUM-201 at 0.8, 1.6 or 3.2 mg/kg per day or recombinant human growth hormone at 34 micrograms/kg per day for six months: annualised height velocity was 6.76, 8.21 and 7.73 cm/yr on the three oral arms against 10.57 cm/yr on injection, and the proportion reaching a positive growth response was 44.4, 72.7 and 72.7 per cent against 100 per cent. A companion report from the same programme described greater growth hormone responses to LUM-201 than to standard secretagogues among subjects of a paediatric trial (PMID 35354138). A phase 3 trial in 150 children began recruiting in May 2026.

Regulatory position is stated plainly in both places it exists. The WADA 2026 Prohibited List names ibutamoren (MK-677) at S2.2.4 — a section titled Growth hormone releasing factors, under which growth hormone secretagogues and their mimetics appear as a bulleted subclass — within the section prohibited at all times rather than in competition only. The FDA describes ibutamoren as an active ingredient not approved by the agency, and has determined it is excluded from the definition of a dietary supplement under section 201(ff)(3)(B)(ii) of the Federal Food, Drug, and Cosmetic Act because it was authorised for investigation as a new drug before any marketing as a supplement. Seven warning letters in the agency's database name it; four were issued in December 2025.

Composition of what circulates has been measured more than once. Van Wagoner and colleagues bought 44 products sold online as selective androgen receptor modulators and analysed them under chain of custody: 17 (39 per cent) contained a different unapproved drug, ibutamoren among them; 4 contained no active compound at all; the label amount matched analysis in only 18. A five-year official medicines control laboratory survey run through the GEON network — 324 samples and 354 results from 14 laboratories in 13 countries, European plus Canada and Australia — listed ibutamoren first among 18 in-scope molecules. One of the December 2025 warning letters concerned two products marketed for children's growth that laboratory analysis found to contain undeclared ibutamoren mesylate.

What is not known

The safety question raised at the end of the commercial programme has never been resolved in public. Adunsky et al. state that the phase 2b hip-fracture trial was stopped early for a congestive heart failure signal in a limited number of patients, but neither the paper's abstract nor the corresponding registry record quantifies it: the registered Merck hip-fracture study lists 83 participants as actual enrolment against the paper's 123 randomised, records no reason for termination and has posted no results, and whether the two documents describe the same protocol cannot be established from public data. No public document located states that this trial ended Merck's development of the compound. The registered Alzheimer's protocol carries a similar mismatch, listing 512 actual participants against the publication's 563 randomised, with no results posted. Of the fourteen registered studies naming this compound, four have posted results. Human pharmacokinetics are effectively unpublished: a validated picogram-level plasma assay exists, but no elimination half-life, absorbed fraction, time to steady state or clearance value for MK-677 was located in any indexed paper. The trial record barely touches the populations that actually take it. The controlled studies located in healthy young adults (PMID 8768828, PMID 9349662, PMID 9467534) each ran seven days and measured growth hormone profiles, sleep and nitrogen balance; no controlled study located ran in healthy young adults for body composition over the periods across which the compound is self-administered. The single published self-administration case report (PMID 36303408) combined MK-677 with LGD-4033 in one 25-year-old man, so its findings on testosterone, liver enzymes and lipids cannot be attributed to either compound. There is no long-term cardiovascular outcome data, no reproductive or developmental data located, and no characterisation of what happens after the two-year mark. Product identity is a further unknown: laboratory surveys keep finding ibutamoren in material sold under other names, and finding declared amounts that do not match analysis.

Questions