MK-677 (Ibutamoren): The Complete Guide

MK-677 — also called ibutamoren or ibutamoren mesylate — is not a peptide. It is a non-peptide ghrelin mimetic: a small molecule developed by Merck in the 1990s that activates the growth hormone secretagogue receptor (GHS-R1a), the same receptor ghrelin and the injectable GHRPs use. Being a small molecule is the entire point — it survives digestion, so it works as a pill while every peptide in this class needs a needle.

It is also, unusually for this category, well studied in humans. Merck ran it through real randomised controlled trials from the mid-1990s onward, including a two-year placebo-controlled trial in older adults. Most compounds people compare it to have nothing remotely like that.

What it never got was an approval. Development ended without MK-677 reaching market for any indication, and the trial record is a large part of why — see the evidence section, because on this compound the trials are the story.

It is not a SARM. Vendors routinely file it under SARMs and the label has stuck, but MK-677 has no interaction with the androgen receptor whatsoever. It acts on the ghrelin receptor and nothing about it is androgenic.

Not FDA approved for any indication; classified as an unapproved drug, and not lawful to sell as a supplement or food. WADA prohibited at all times under S2 (includes secretagogues and mimetics). Half-life is roughly 24 hours.

This is the section that matters. MK-677 has a fuller human trial record than almost anything else in this category, and the results are strikingly consistent across studies, populations and durations. They are also not what the marketing says.

It raises GH and IGF-1. This part is solid.

Chapman 1996 (PMID 8954023), 32 healthy adults aged 64–81 at 2, 10 or 25 mg daily: the 25 mg dose raised mean 24-hour GH by 97 ± 23% and lifted IGF-1 from 141 to 219 mcg/L at two weeks and 265 mcg/L at four — into the normal range for young adults. Every subsequent trial reproduced the effect. If the question is "does it work on the axis", the answer is yes, reliably.

It did not reduce fat.

This is where it starts diverging from the pitch. In the two-year trial (Nass 2008, PMID 18981485, 65 healthy adults aged 60–81), abdominal visceral fat and total fat mass were primary endpoints — and neither changed significantly. Limb fat actually rose more on MK-677 than on placebo (1.1 kg vs 0.24 kg, p = 0.001), and body weight climbed 2.7 kg versus 0.8 kg. Svensson 1998 (PMID 9467542), in obese men, found the same thing: fat-free mass up, total and visceral fat not significantly changed.

Fat-free mass did increase in Nass — +1.1 kg against −0.5 kg on placebo (p < 0.001), a real result over two years. But gaining 1.1 kg of lean mass and 1.1 kg of limb fat is not body recomposition. It is weight gain with the ratio preserved.

It did not improve strength or function. Three times.

Nass states it flatly: "Increased fat-free mass did not result in changes in strength or function." Isokinetic strength, physical function and quality of life were all measured, over two years, and none moved.

Two independent hip-fracture trials found the same. Bach 2004 (PMID 15066065), 161 patients: IGF-1 rose 84% versus 17% on placebo, and there was no significant improvement in functional performance or the overall sickness-impact score. Adunsky 2011 (PMID 21067829), 123 patients: IGF-1 up 51.4 ng/mL, gait speed improved, and most other functional measures did not.

Three trials, three populations, one pattern — the biomarker moves and the outcome does not. That gap is the single most important thing to understand about this compound.

The trial that was stopped early.

Adunsky 2011 was terminated before completion because of a congestive heart failure safety signal. The authors' conclusion is unambiguous: "MK-0677 has an unfavorable safety profile in this patient population."

Read the qualifier honestly in both directions. That population was elderly patients recovering from hip fracture, which is not a healthy 35-year-old, so the risk does not transfer directly. But it is also the only trial that ran MK-677 in a frail population and looked hard at cardiac outcomes, and it was stopped.

The metabolic cost is consistent.

Chapman saw fasting glucose rise from 5.4 to 6.8 mmol/L at four weeks (p < 0.01). Nass saw fasting glucose rise and insulin sensitivity decrease over two years. This is not an idiosyncratic reaction — it is the expected consequence of sustained GH elevation, and it showed up in the short trials and the long one alike.

Where it genuinely delivered: sleep and nitrogen balance.

Sleep. Copinschi 1997 (PMID 9349662) is a placebo-controlled crossover trial with sleep architecture as the endpoint. In young subjects, 25 mg raised stage IV sleep duration by roughly 50% and REM by over 20%, and deviations from normal sleep fell from 42% to 8%. In older subjects, REM rose nearly 50% with shortened REM latency. Of everything claimed for MK-677, the sleep benefit has the most direct trial support — which is worth saying, because it is often listed as anecdotal.

Catabolic protection. Murphy 1998 (PMID 9467534) calorie-restricted 8 healthy volunteers to 18 kcal/kg/day. On placebo, nitrogen balance was −1.48 g/day; on MK-677 it was +0.31 g/day (p < 0.01). A small, short study, but a clean demonstration of nitrogen retention under a deficit.

Bone. Svensson 1998 (PMID 9661080) found increased markers of both bone formation and resorption in obese young men — increased remodelling, which Nass also observed. Higher turnover is not the same as higher density, and no trial has shown a density benefit.

On cortisol, the picture is not what you have read

MK-677 is routinely described as cortisol-raising. The trials disagree with each other. Chapman found circulating cortisol did not change. Svensson found serum and urinary cortisol not increased at 2 or 8 weeks. Only the two-year Nass trial found a rise, of 47 nmol/L (p = 0.020) — about 1.7 mcg/dL. Prolactin is firmer: Chapman measured a 23% increase, though it stayed inside the normal range.

So: a modest prolactin rise is well supported, and a cortisol rise appears with long-term use but not in the shorter trials. "MK-677 spikes cortisol" overstates a mixed record.

What was actually studied

Unusually for this category, the practical dose and the studied dose are the same number. 25 mg once daily is what nearly every trial used — Chapman (alongside 2 and 10 mg), Svensson, Murphy, Nass for two years, and both hip-fracture trials. Chapman's dose-ranging found 25 mg produced the largest GH and IGF-1 response.

So the dosing below is not extrapolation. What is extrapolation is the goal: the trials were run in older adults, obese men, calorie-restricted volunteers and hip-fracture patients — not in healthy trained adults seeking body composition change.

Standard protocol

Dose	10–25 mg once daily
Route	Oral — no reconstitution, no refrigeration
Timing	Before bed, which aligns with the sleep findings and keeps the appetite effect overnight
Titration	10 mg for the first week or two, then 25 mg if appetite and fluid retention are tolerable
Duration	Nass dosed continuously for two years; there is no efficacy argument for cycling

On cycling

No trial shows the IGF-1 elevation fading, so cycling is not needed to preserve the effect — the two-year data settles that. The argument for breaks is metabolic, not pharmacological: fasting glucose rises and insulin sensitivity falls with sustained use, and time off lets those normalise. That is a different reason from the receptor-desensitisation logic used with the injectable GHRPs.

Monitoring is not optional here

Given a two-year trial showing reduced insulin sensitivity, glucose monitoring is the minimum: fasting glucose and HbA1c at baseline, around four weeks, then periodically. IGF-1 confirms the compound is working and that levels are not being driven high — and note that higher IGF-1 is not automatically better, since the mortality relationship is U-shaped (see the HGH guide). A baseline prolactin is reasonable for anyone already near the top of the range.

Documented in trials. Nass reported the most frequent effects as increased appetite that subsided after a few months, plus transient mild lower-extremity oedema and muscle pain. Across studies: rising fasting glucose and falling insulin sensitivity, a 23% prolactin increase (Chapman), a 47 nmol/L cortisol increase over two years (Nass, but not seen in the shorter trials), and increased bone remodelling.

Appetite is the mechanism, not a side effect. MK-677 is a ghrelin mimetic; hunger is the receptor doing its job. With a 24-hour half-life there is no window in which it is absent. For anyone in a deficit this is the effect most likely to end the experiment.

Commonly reported outside the trials (lower evidentiary weight): fluid retention in the face and hands, daytime lethargy especially with morning dosing, transient tingling or numbness in the hands in the first weeks — consistent with the carpal-tunnel-type symptoms seen with GH elevation generally — joint stiffness and vivid dreams.

The cardiac signal. The hip-fracture trial (PMID 21067829) was stopped early for congestive heart failure and concluded the safety profile was unfavourable in that population. It is the most serious finding in the MK-677 literature and it belongs in view even though the population was frail and elderly.

Do not use with:

• Diabetes, pre-diabetes or insulin resistance — this is the standout contraindication for MK-677 specifically. Raising fasting glucose and lowering insulin sensitivity is a replicated finding, not a theoretical risk.
• Active malignancy, or a history of one without oncology input — standard for anything driving the GH/IGF-1 axis.
• Congestive heart failure or significant cardiovascular disease, on the basis of the terminated trial.
• Pregnancy or breastfeeding — no safety data.

Use caution with: metabolic syndrome, already-elevated prolactin, uncontrolled hypertension, a history of carpal tunnel syndrome, and liver or kidney impairment.

Interactions: insulin and oral diabetes medications — glucose management will need revisiting. Running MK-677 alongside a GLP-1 agonist is pharmacologically incoherent: one drives hunger through the ghrelin receptor while the other suppresses it, and the insulin-sensitivity cost of MK-677 works directly against what the GLP-1 is for.

Athletes: WADA prohibits MK-677 at all times under S2, which explicitly covers secretagogues and mimetics. It is also banned across the major US sporting bodies.

MK-677 and ipamorelin hit the same receptor. The differences are selectivity, kinetics, and how much is known.

	MK-677	Ipamorelin
Route	Oral	Subcutaneous injection
Half-life	~24 h — sustained exposure	~2 h — a pulse that clears
Appetite	Strong and continuous	Minimal
Prolactin	+23% (Chapman)	No rise at 200× the GH dose, in swine
Glucose / insulin	Worsens, replicated	Not established either way
Human evidence	Multiple RCTs including 2 years	One PK study, one failed Phase 2

The honest comparison cuts both ways, and most write-ups only give one side of it. MK-677's side-effect profile is genuinely worse — the appetite is relentless, the metabolic cost is real and replicated. But MK-677 has by far the longest human exposure of any GH secretagogue — two years — and that data is what tells us the fat and function endpoints did not move. Ipamorelin’s longest published human exposure is seven days; tesamorelin has 20-week and longer trials but is a GHRH analogue rather than a ghrelin agonist. Ipamorelin looks cleaner partly because nobody has run the two-year trial that would find its problems.

Preferring ipamorelin on side-effect grounds is defensible. Preferring it because its results are better is not something the evidence supports, since those results have never been measured.

Also worth separating: MK-677 is not a SARM despite where it gets shelved, and it is not related to HGH — it asks your pituitary for GH rather than supplying it, so the feedback loop stays intact and there is a ceiling on the response.