Peptides for perimenopause weight gain: what actually has evidence

Somewhere between 43 and 50, a lot of women notice the same thing. The food is the same. The walking is the same. The weight moves anyway, and it moves to the middle. Then you land in the peptide internet, where someone is telling you a $90 vial of AOD-9604 is what your metabolism has been waiting for. This is that conversation with the numbers left in: what was studied, in whom, what it costs, and where the honest answer is that nobody has tested this.

What actually changes, and why it isn't willpower

What changes in perimenopause is less how much fat you carry than where. As oestrogen becomes erratic and then falls, storage shifts from hips and thighs toward the abdomen, and more of it becomes visceral — packed around the organs rather than under the skin. That is the fat associated with insulin resistance and cardiometabolic risk, which is why it matters beyond your jeans.

Two things happen alongside it. Lean muscle declines with age, lowering resting energy expenditure; and sleep gets worse, which independently affects appetite regulation and next-day food choices. So a woman in perimenopause is often eating slightly more, moving less, carrying less muscle, and storing what she eats somewhere more metabolically active. None of that is a character flaw. It is also why "peptides for menopause belly fat" is such effective copy: it names a real change. The question is whether the products attached to it do anything about it.

The only randomised perimenopause data in the category

In 2025, Tchang and colleagues published a post-hoc analysis in Obesity pooling 2,542 women from the SURMOUNT-1, -3 and -4 tirzepatide trials, staged retrospectively as pre-, peri- or postmenopausal. Supported

Those are large numbers, and almost nothing else on this site produces numbers like them. The perimenopausal group did about as well as the postmenopausal group and slightly less well than the premenopausal one. Reproductive stage did not appear to blunt the drug much.

Now the caveats, in full

This was post hoc. Women were not enrolled by menopausal stage — the staging was applied afterwards to existing data, which makes the finding hypothesis-generating rather than confirmatory. It was also funded and authored by Eli Lilly, which makes tirzepatide; several authors are company employees.

That does not make the numbers wrong, and the underlying trials were randomised and placebo-controlled — a genuinely higher standard than anything else here can offer. But you should know who paid for the most quotable statistic in the field. Nobody has run a trial designed to test tirzepatide in perimenopause.

Compare the options

Two of these columns — legal status and cost — are missing from every other page ranking for this query.

Research-vial prices are what vendor listings advertise, not what the vial contains — see the 30% problem. Brand prices are the manufacturers' own self-pay programmes as published in September 2026.

Semaglutide, and the analysis nobody ran

Semaglutide has enormous randomised evidence for weight reduction. STEP-1 randomised 1,961 adults and reported a mean 14.9% reduction at 68 weeks versus 2.4% on placebo, with 50.5% losing 15% or more, in a majority-female population. Supported

What does not exist is the semaglutide equivalent of the Tchang analysis. Nobody has staged the STEP programme by menopausal status. So anyone telling you semaglutide is better or worse than tirzepatide "in perimenopause" is comparing a drug analysed that way against one that has not been, and there is no head-to-head trial in perimenopausal women either.

Does hormone therapy change what a GLP-1 does?

A Mayo Clinic retrospective cohort of 120 postmenopausal women on tirzepatide for at least 12 months, published in The Lancet Obstetrics, Gynaecology & Women's Health, found those also using menopausal hormone therapy lost about 35% more weight than those on tirzepatide alone. Mixed

The authors were more careful with it than most of the coverage. It was not randomised, and their stated caution is that they cannot say hormone therapy caused the difference. Women who stay on hormone therapy may already be more engaged with their health; or symptom relief may have improved sleep enough to sustain diet and activity changes. Both fit the data as well as a drug effect does.

If you are already on hormone therapy and considering a GLP-1, nothing here argues against it. Starting hormone therapy specifically to amplify one is a different question, and a 120-woman retrospective cohort is a thin reason. More in peptides and HRT.

"Microdosing" a GLP-1: the word means nothing

Microdosing is the most common thing being sold to perimenopausal women right now, framed as gentler, more sustainable, less muscle loss, fewer side effects. There is no pharmacological definition of a GLP-1 microdose. No dose-finding study established one, and no randomised trial of any microdosing protocol exists. The word has no agreed clinical meaning, which is why one clinic's "microdose" is a different quantity from the next one's.

What is known is that the effect is dose-dependent — SURMOUNT and STEP escalated doses and saw larger effects at higher doses, which is the entire basis of the titration schedules on the approved labels. A sub-therapeutic dose should produce a sub-therapeutic result. That is arithmetic, not criticism, and a smaller dose may be what someone wants if the goal is tolerability. But it is sold as a smarter version of the drug rather than a smaller one. Nobody has tested whether it preserves more lean mass, or works differently in perimenopause. Nobody has tested it at all.

What you lose besides fat

The SURMOUNT-1 body-composition substudy used DXA in 160 participants at baseline and week 72. Body weight fell 21.3%, fat mass 33.9%, lean mass 10.9% — roughly 75% of the weight lost was fat and 25% lean tissue.

That is not evidence the drug is uniquely catabolic — similar ratios appear with diet-induced loss. It is a reason to count resistance training and protein as part of the protocol, not extras. See muscle and bone.

And bone — where the evidence is better than the fear

The worry is reasonable in principle: bone density falls during significant weight loss, and perimenopausal women are already in the steepest phase of bone loss they will experience. The direct evidence does not support alarm. A 2025 meta-analysis of 25 randomised trials in type 2 diabetes found no significant effect of GLP-1 receptor agonists on fracture risk, and improvements rather than declines in bone density at the lumbar spine, femoral neck and total hip. Mixed

Two limits: those trials were in people with type 2 diabetes, and were not designed as bone studies. Fair summary: no fracture signal in the randomised data we have, a real general concern about rapid weight loss and bone, and no bone density trial in perimenopausal women on a GLP-1. If you have osteopenia or a fracture history, ask someone who can order a DXA.

The interaction nobody discloses

If you take one thing from this page, take this, because as far as we can tell no peptide clinic protocol and no product label mentions it. Every growth-hormone secretagogue sold for menopause belly fat — sermorelin, CJC-1295, ipamorelin, tesamorelin — works by increasing growth hormone release, the effect running through IGF-1 and hepatic fat oxidation. That is the mechanism you are paying for.

Oral oestrogen impairs exactly that mechanism. This is established endocrinology, not speculation. Leung, Johannsson, Leong and Ho reviewed it in Endocrine Reviews in 2004: oral, but not transdermal, oestrogen impairs the metabolic action of growth hormone in the liver, causing a fall in IGF-1 production and fat oxidation. Their phrase for the consequence in postmenopausal women is worth quoting exactly — it "results in a loss of lean tissue and a gain of body fat." The proposed mechanism is induction of SOCS-2, which inhibits growth hormone's JAK/STAT signalling. IGF-1 generation testing agrees, and found transdermal reduced the response less.

This does not apply to GLP-1s, which do not work through the growth hormone axis. They have a different interaction: slowed gastric emptying can alter absorption of oral estradiol and progesterone. Route matters either way — see can you take peptides with HRT.

Tesamorelin, sermorelin, CJC-1295, ipamorelin: what was actually studied

Tesamorelin — a real drug, for someone else

Tesamorelin is the strongest case in this group and still not a case for you. It is FDA-approved as EGRIFTA, with the WR formulation approved in 2025, for one indication: excess abdominal fat in HIV-infected adults with lipodystrophy. Pivotal trials randomised 273 and 270 patients against 137 and 126 on placebo; visceral adipose tissue fell 18% and 14% versus a 2% increase. A genuine, replicated visceral fat effect. Thin for this population

What the marketing leaves out. The label's Limitations of Use state that long-term cardiovascular safety is not established and that it is not indicated for weight loss — it is weight-neutral and redistributes fat. The trials were about 85% male, mean age 48; no perimenopausal data. HbA1c reached 6.5% or higher in 5% of treated patients versus 1% on placebo, and at 26 weeks IGF-1 exceeded 2 standard deviations in 47%. It is contraindicated in active malignancy, disrupted hypothalamic-pituitary axis, and pregnancy. And visceral fat returns when you stop.

And for anyone told a growth hormone peptide helps brain fog: a 2025 Phase 2 trial in 73 people with HIV and abdominal obesity cut waist circumference and raised IGF-1 by 53%, with no significant neurocognitive benefit (p = .573). See energy and brain fog.

Sermorelin — the "FDA-approved" sleight of hand

You will read that sermorelin is FDA-approved. It was, as GEREF, in 1990 and 1997 — discontinued commercially in 2008, approval withdrawn in 2009. FDA determined in 2013 that the withdrawal was not for reasons of safety or effectiveness, a real distinction, but still a commercial decision.

Because it was a component of an approved drug it satisfies one of the three 503A conditions, so it can lawfully be compounded — which is why sermorelin is the peptide your local clinic can actually prescribe. That is a legal fact, not an efficacy one. GEREF's approved indication was as a diagnostic agent and for paediatric growth hormone deficiency: sermorelin contains a substance formerly approved for a different indication in a different population. Nobody has tested it here. No human data

CJC-1295 and ipamorelin — two problems the seller won't mention

The molecule studied is often not the molecule sold. The single human study cited for CJC-1295 (Teichman et al., JCEM 2006) tested CJC-1295 with DAC, half-life several days. Much of what is dispensed is CJC-1295 without DAC — half-life around 30 minutes, no published human trials at all. From a vial, most buyers cannot tell which they have.

The only registered efficacy trial was terminated. NCT00267527, a Phase 2 study of CJC-1295 in 120 patients with HIV-associated visceral obesity, was terminated with no reason posted and never published. It is the only efficacy trial CJC-1295 has had.

Ipamorelin is the clearest case here of a female protocol written from male data. Its only human pharmacokinetic study — Gobburu et al., 1999 — used eight healthy male subjects at each of five dose levels. Every half-life, clearance and potency figure in circulation comes from those men. That study dosed in nmol per kilogram; clinics dispense flat doses in micrograms, so a 55 kg woman and a 95 kg man on "250 mcg ipamorelin" get roughly a 1.7-fold difference in exposure nobody accounts for.

There is no human trial of CJC-1295 or ipamorelin, alone or combined, for body composition, fat loss, muscle gain or sleep, in either sex. The muscle finding you have seen quoted is a mouse study. No human data

AOD-9604: a drug that failed, sold as one that works

AOD-9604 is a fragment of human growth hormone marketed almost exclusively for fat loss, and the cleanest example in the category of a negative result made invisible. It has zero clinical trials indexed on PubMed and zero registered studies on ClinicalTrials.gov — we checked both directly. Its Phase 2b obesity programme, in roughly 500 subjects, failed to separate from placebo and was abandoned. That failure was documented in company communications and never peer-reviewed: the trial that would tell you whether it works was run, then disappeared.

Some marketing leans on an Australian food-ingredient safety determination for an oral form — a judgment about not causing harm as a food ingredient, not a finding of efficacy, and not applicable to injection. No published human data

The legal picture, and the error nearly every site is making

Compounded GLP-1s

Compounded semaglutide and tirzepatide were widely available because both drugs were on FDA's shortage list and FDA exercised enforcement discretion. That is over: it ended for tirzepatide on 18 February 2025 (503A pharmacies) and 19 March 2025 (503B outsourcing facilities), and for semaglutide on 22 April and 22 May 2025. Neither drug is on the shortage list, and compounded copies are now generally impermissible.

The exception compounders use to keep selling is "personalised dosing" — that a customised dose is not a copy. That position is legally contested, and it is why your telehealth microdosing subscription still exists.

The peptides — and the sentence to stop believing

In April 2026 FDA removed BPC-157, TB-500, Semax, Epitalon, MOTS-c, GHK-Cu and others from Category 2, the list flagged as presenting significant safety risks. In July 2026 the Pharmacy Compounding Advisory Committee voted on seven peptides. FDA's own career scientists recommended against all seven; the committee voted in favour of six anyway, mostly narrowly, and rejected emideltide (DSIP).

The 503A test explains the map: a compounder may use a bulk substance only if it has a USP or NF monograph, is a component of an approved drug, or is on the 503A Bulks List. BPC-157, AOD-9604, CJC-1295 and ipamorelin pass on none of the three.

How to decide, and the 30% problem

Independent laboratory testing by BioTools has found problems in nearly 30% of the peptide vials it tested — mislabelled, underdosed, overdosed, or contaminated. Co-founder Rina Dukor: among the thousands of tests they run they find live or dead bacteria in a lot of the vials, and that could lead to hospitalisations.

Which reframes the question. Before asking "does this peptide work," ask "is this vial the thing on the label." If you are going grey-market because the prescription route is unaffordable, batch certificates of analysis matter more, not less. Also worth being honest with yourself about:

• Perimenopausal women can still conceive. GLP-1s are not recommended in pregnancy, tesamorelin is contraindicated in it, and irregular cycles are not contraception.
• Breast cancer, a BRCA variant, tamoxifen or an aromatase inhibitor: growth hormone secretagogues raise IGF-1 by design. Ask your oncology team.
• Tirzepatide and semaglutide are contraindicated with a personal or family history of medullary thyroid carcinoma or MEN2.
• Thyroid disease is common here. If yours is untreated or unstable, that comes first.
• Have a stopping plan before you start. Weight regain after GLP-1 discontinuation is well documented, and tesamorelin's visceral fat effect does not persist after stopping. Nobody sells the exit.

What nobody has tested

The list is longer than the list of what has been tested. Nobody has run a GLP-1 trial designed prospectively in perimenopausal women, or staged the semaglutide trials by menopausal status, or tested whether "microdosing" preserves lean mass, or run a bone density trial in perimenopausal women on a GLP-1. Nobody has tested tesamorelin in perimenopausal women, or sermorelin, CJC-1295 or ipamorelin for body composition in any human of either sex. Nobody has published the AOD-9604 trial that failed. And nobody has tested any growth hormone secretagogue in women stratified by whether their oestrogen is oral or transdermal — the one interaction we can predict from established endocrinology.

The pattern is consistent: protocols are male-derived, doses are flat where the source pharmacology was weight-normalised, and negative results are systematically invisible. CJC-1295's only efficacy trial was terminated with no reason posted; two ipamorelin Phase 2 studies never posted results; AOD-9604's failed Phase 2b was never published.

None of which means nothing works. Tirzepatide clearly does something substantial in thousands of women, perimenopausal ones included. It is a reason to be suspicious of anyone selling confidence about the rest.

Questions people actually ask

If weight is one of several things that shifted at once, the rest of this series covers energy and brain fog, sleep, muscle and bone, skin, hair and collagen, and libido and mood — same rule throughout.