Peptides for muscle loss and bone density in perimenopause

Most writing about perimenopause covers the loud symptoms — heat, sleep, mood, weight. Muscle and bone are quiet. Nothing hurts, nothing wakes you at 3am, and then a wrist fracture from a low fall turns into a conversation about your spine, and you find out the thing nobody mentioned was the thing with the longest consequences.

It is also where peptide advertising is most confident and least supported. Growth hormone secretagogues — sermorelin, CJC-1295, ipamorelin, tesamorelin — are sold to women in their forties and fifties for lean mass and body composition; BPC-157 and TB-500 for joints and tendons. What follows is what has actually been measured, in whom, and what happened. In several cases the honest answer is that nobody has tested this in anyone, let alone in a woman going through the menopause transition.

What actually changes, and how much

Two things are happening, and not at the same scale.

Bone: the fast, well-documented loss

The Study of Women's Health Across the Nation followed 862 women through their final menstrual period with repeated bone scans. Loss began about a year before that period and decelerated — but did not stop — roughly two years after. Across ten years women lost 10.6% of lumbar spine bone mineral density and 9.1% at the femoral neck, most of it in the three-year window straddling the final period: 7.4% at the spine, 5.8% at the hip.

Lean mass: real, but smaller than you've been told

Here the honest number is less dramatic, and worth knowing before you spend money. SWAN also measured body composition by DXA in 1,246 women. Lean mass was rising slightly before the transition (about 0.2% a year); during it, that reversed to a fall of about 0.2% a year — roughly 60 grams annually, an average total loss of about 0.5%, or 0.2 kg, across the entire transition. Afterwards the slope flattened to zero.

The visible change comes from fat: the rate of fat gain roughly doubled at the start of the transition, from 0.25 kg to 0.45 kg a year. So the proportion of you that is lean fell about 1.9% — what you see in the mirror — while the absolute muscle lost was small.

That matters commercially. A page selling a peptide for "menopausal muscle loss" is describing a change averaging about two hundred grams. The strength and function losses women report are real, but they are driven more by inactivity, under-eating protein, disrupted sleep and the fat-to-lean shift than by any collapse in muscle tissue — which is good news, because those are all things you can act on.

Every dose number you've been given came from men

Ipamorelin is the purest example of a problem running through this whole category.

There is exactly one human pharmacokinetic and pharmacodynamic study of ipamorelin, published in 1999. It used five infusion rates with eight healthy male subjects at each dose level. Every circulating figure you will read about ipamorelin — the two-hour terminal half-life, the clearance of 0.078 L/h/kg, the SC50 of 214 nmol/L for half-maximal GH stimulation — comes from those men.

Now look at how it is sold. Clinic protocols and vendor pages give flat doses: "200–300 mcg", "250–500 mcg". The source pharmacology was weight-normalised — the 1999 study dosed in nmol per kilogram. A 55 kg woman and a 95 kg man taking the same 300 mcg receive roughly a 1.7-fold different exposure per kilogram, and nothing in any protocol accounts for it.

Nor is this a grievance dressed up as pharmacology. GH secretion, IGF-1 response and tissue sensitivity to GH all differ by sex; in treating growth hormone deficiency, women generally need higher GH doses than men to reach the same IGF-1 level. Body composition shifts it further: in IGF-1 generation testing with a fixed, non-weight-based GH bolus, people with more body fat produced larger IGF-1 increases than lean people on the identical dose. A flat dose, derived from weight-normalised data generated in young men, given to women across a 40 kg weight range, with no dose-finding study in women at all — that is the protocol.

The HRT interaction nobody on a peptide label mentions

Oral estrogen — but not transdermal — impairs growth hormone's metabolic action in the liver. The mechanism is worked out: estrogen arriving via the portal circulation induces SOCS-2, a suppressor protein that inhibits GH's JAK/STAT signalling. IGF-1 production falls. Fat oxidation falls. And the documented downstream consequence in postmenopausal women, in the words of the definitive Endocrine Reviews paper on the subject, is "a loss of lean tissue and a gain of body fat."

Three qualifications, because this deserves precision rather than drama. Route is the variable, not hormone therapy. It is a first-pass liver effect; transdermal estradiol largely bypasses it, though one IGF-1 generation study found transdermal reduced GH responsiveness somewhat, clearly less than oral. It is not an argument against HRT — hormone therapy has far better bone evidence than any peptide here, so if anything it argues against the peptide. And it is not a reason to change your prescription alone: estrogen route is chosen for clotting risk, migraine history and absorption, which matter more than this. We go through the detail on peptides and HRT.

GH secretagogues: what has actually been tested

CJC-1295 and ipamorelin No human data

The molecule studied is not always the molecule sold. The one substantial human study of CJC-1295 (2006, healthy adults) tested CJC-1295 with DAC, half-life 5.8 to 8.1 days. Much of what clinics and vendors dispense is CJC-1295 without DAC — modified GRF 1-29, half-life about thirty minutes, no published human trials whatsoever. From a label, most buyers cannot tell which they have.

The only efficacy trial was terminated. CJC-1295's sole registered efficacy study, a Phase 2 in HIV-associated visceral obesity with 120 participants, was terminated in 2006. No reason posted. Never published. That is the entire registered efficacy record.

The completed ipamorelin trials were never reported. Two Phase 2 trials in post-operative ileus finished — 117 participants in 2009, 320 in 2014. Neither posted results. In drug development, a completed Phase 2 that is quietly never published and never followed up is almost always one that did not work. Ipamorelin was not pursued.

As for muscle: there is no human trial of CJC-1295 with ipamorelin for body composition, fat loss or muscle gain in either sex. The finding that gets cited — improved maximum tetanic tension in glucocorticoid-induced muscle loss — is a mouse study, and a 2026 review in the American Journal of Sports Medicine says so plainly.

Sermorelin Thin

Sermorelin is where a legal fact gets sold as an efficacy fact. Approved as GEREF in 1990, discontinued commercially in 2008, approval withdrawn in 2009 — and the FDA determined in 2013 the withdrawal was not for safety or effectiveness reasons. That is why sermorelin can be lawfully compounded and BPC-157 cannot: a regulatory pathway, not a finding that it works. What GEREF was approved for matters too — a diagnostic agent and paediatric growth hormone deficiency, not adult body composition, not bone, not perimenopause. Accurately stated, sermorelin contains a substance formerly approved for a different indication in a different population. There are no modern trials of it for muscle or bone in midlife women.

Tesamorelin Mixed

Tesamorelin has a genuine current FDA approval, and reading it carefully is instructive. It is approved only for reduction of excess abdominal fat in HIV-associated lipodystrophy, and its label carries an explicit Limitation of Use: not indicated for weight loss. It redistributes fat rather than reducing weight, and long-term cardiovascular safety is not established.

The pivotal trials ran 543 participants on drug against 263 on placebo over 26 weeks; visceral fat fell 14–18%. They were roughly 85% male, mean age 48, with no perimenopausal data, and the effect reverses on discontinuation. HbA1c reached 6.5% or higher in 5% of treated participants versus 1% on placebo — hazard ratio 3.3. IGF-1 exceeded two standard deviations above normal in 47% and three SDS in 36% at 26 weeks. It is contraindicated in active malignancy, disrupted hypothalamic-pituitary axis, and pregnancy — which matters, because perimenopausal women can still conceive. Brand pricing runs around $37,500 a year.

Cost figures are approximate US market ranges observed in September 2026 and vary enormously by vendor, vial size and dose. Research-vial pricing and clinic pricing differ by an order of magnitude, and neither price is a quality signal.

BPC-157 and TB-500 for joints and tendons

BPC-157 No human data

There are around 230 papers on BPC-157 in PubMed. Zero are indexed as clinical trials. A 2025 narrative review from the University of Utah, with no industry conflicts, found only three pilot studies have examined BPC-157 in humans at all — intra-articular knee pain, interstitial cystitis, and an intravenous safety report — and concluded it remains investigational.

That safety report is routinely cited as "the human safety study." It enrolled two people: a 58-year-old man and a 68-year-old woman, at a private Florida clinic, both already receiving intravenous BPC-157 before the study began.

The one registered Phase 1 trial, n=42, completed in March 2016; its status is still listed as unknown and results have never been posted, a decade on. The trial currently recruiting — BPC-157 for acute hamstring strain, n=120 — caps enrolment at age 45, structurally excluding most of perimenopause from the only ongoing human study.

The FDA's own advisory-committee briefing document was blunt: lack of evidence to support effectiveness; quality data lacking or inadequate; immunogenicity risk potentially amplified by aggregation; and no studies at all of the oral, subcutaneous, nasal or transdermal routes actually being sold. Gut, joint and tendon claims rest on rat studies. Oral BPC-157, the form sold most heavily, has never been studied in a human being.

TB-500 No human data

TB-500's problem is subtler. Thymosin beta-4 is a real, naturally occurring 43-amino-acid peptide with real human trials: a Phase 2 topical study in venous stasis ulcers across eight European sites, an intravenous safety study in 40 healthy volunteers, a first-in-human recombinant study in China.

TB-500 is not that molecule. It is a short synthetic fragment, essentially the seven-residue actin-binding domain, and data generated with the full-length peptide do not transfer to a fragment — even before you notice those trials were in venous leg ulcers and cardiac ischaemia, not tendon, not joint, not any injury-recovery indication TB-500 is sold for. The 2026 AJSM verdict on both: "human orthopaedic data are lacking." Both are WADA-banned.

What actually has the evidence: load, protein, hormones

This is the part a peptide site is not supposed to write, so here it is with the numbers.

Heavy resistance training Supported

The LIFTMOR trial randomised 101 postmenopausal women with low bone mass (T-score below −1.0, mean age 65) to eight months of supervised high-intensity resistance and impact training — twice weekly, thirty minutes, five sets of five reps above 85% of one-rep maximum — or a home-based low-intensity programme. Lumbar spine bone mineral density rose 2.9% in the training group and fell 1.2% in controls; femoral neck density rose 0.3% against a 1.9% fall. Cortical thickness, height and every functional measure improved significantly. Compliance was 92%, with one adverse event in the whole trial: a minor back spasm.

Set that against the bone-loss figures at the top of this page. Twice a week, half an hour, and the direction of travel reverses. Nothing in a vial on this page has produced a bone density number in a woman at all.

Protein Supported

The PROT-AGE consensus recommends at least 1.0–1.2 g of protein per kg of body weight daily for older adults to maintain and regain lean mass and function, and 1.2 g/kg or more for those exercising — for a 68 kg woman, roughly 70–85 g a day as a floor. Most women in this age group eat considerably less.

The caveat cuts against supplement marketing: a 2022 meta-analysis of whey protein in postmenopausal women found it improved strength and lower-limb lean mass only in combination with resistance training. Without training, no significant effect on either. Protein makes the training work; it does not substitute for it.

Hormone therapy Supported for bone

In the Women's Health Initiative, 16,608 postmenopausal women were randomised to conjugated equine estrogen plus medroxyprogesterone or placebo over an average of 5.6 years. Fractures occurred in 8.6% of the hormone group and 11.1% of placebo — hazard ratio 0.76 — and the effect held across every subgroup, including women not at high fracture risk. Total hip bone density rose 3.7% at three years against 0.14% on placebo.

The WHI also found no net benefit on its global risk-benefit index in that older cohort, which is why the framing changed afterwards. The current Menopause Society position is that for women under 60, or within ten years of menopause onset, with no contraindications, the benefit-risk ratio is favourable for vasomotor symptoms and prevention of bone loss — a much closer description of most readers here than of the average WHI participant. For muscle specifically the effect is modest, and honesty requires saying so: a meta-analysis of 23 studies found hormone therapy associated with roughly 5% greater muscle strength.

Creatine earns its row: a 12-month trial randomised 47 postmenopausal women to creatine or placebo alongside supervised resistance training, and femoral neck density fell 1.2% on creatine versus 3.9% on placebo. Small trial, substantial dropout (33 of 47 completed) — promising rather than settled, at about ten dollars a month.

Collagen peptides and bone density

Collagen peptides sit oddly here: they are the one thing on the shelf actually trialled in the right people. A 12-month randomised, placebo-controlled, double-blind study gave 5 g of specific collagen peptides or placebo daily to 131 postmenopausal women with age-related low bone density (mean age 64, mean spine T-score −2.4). Spine T-score rose 0.1 on treatment and fell 0.03 in controls; femoral neck rose 0.09 against a 0.01 fall. Formation marker P1NP rose on treatment, resorption marker CTX-1 rose in controls. An open-label follow-up of 31 women to four years reported continued increases.

What honesty requires you to notice: 131 women is modest, 102 completed, the effect is reported as a change in T-score rather than a percentage of density, and 0.1 of a T-score is a small absolute move. The study was partly financially supported by GELITA AG, which makes the collagen peptide used, and a co-author heads the Collagen Research Institute — though the published conflict-of-interest statement declares none. The four-year follow-up was open-label, 31 participants, no control group: closer to an observation than a trial.

That does not make it worthless. It makes it one industry-supported year-long RCT in exactly the right population, showing a small favourable signal, never independently replicated at scale — still more human evidence here than every injectable peptide on this page combined. We look at collagen from the skin side in the skin, hair and collagen guide.

If you're losing weight on a GLP-1

In the SURMOUNT-1 body composition substudy, 160 participants — 73% of them women — had DXA scans at baseline and week 72. Body weight fell 21.3% on tirzepatide, fat mass 33.9%, and lean mass 10.9%. Of the total weight lost, roughly 75% was fat and 25% was lean, and that proportion held across sex, age and weight-loss tertile.

What is unusual is the magnitude, arriving in a woman whose bone is already in its fastest-losing window and whose lean mass has just turned the corner. Rapid weight loss also raises bone resorption markers, though randomised data has not so far shown measured density change over trial durations. The response is not a GH secretagogue, which has never been tested for this purpose in anyone — it is resistance training and protein maintained through the weight loss, plus a conversation with your prescriber about pace. We cover the GLP-1 evidence properly in the weight guide.

If you've had breast cancer, or carry a BRCA variant

No other site ranking for these terms addresses this, and it is straightforward to state carefully.

GH secretagogues work by raising GH and, downstream, IGF-1 — that is the point of them. In the tesamorelin pivotal trials, IGF-1 exceeded three standard deviations above the normal range in 36% of participants at 26 weeks. Separately, circulating IGF-1 is positively associated with breast cancer risk in large prospective cohorts: in UK Biobank, across 394,388 cancer-free participants, each 5 nmol/L higher IGF-1 concentration carried a hazard ratio of 1.11 for breast cancer. That is an observational association in people whose IGF-1 was whatever it naturally was — not an experiment in which anyone's IGF-1 was raised deliberately.

It cuts the other way too: aromatase inhibitors accelerate bone loss substantially, and survivors are often the women most in need of something for bone. That need is real — and what has been studied in that setting is bisphosphonates, denosumab, resistance training and oncology-supervised care, not an untested secretagogue.

Cost, legality, and what's in the vial

Legality. In April 2026 the FDA removed BPC-157, TB-500, MOTS-c, GHK-Cu and others from Category 2, its list of substances that may present significant safety risks. In July 2026 an advisory committee voted in favour of six of the seven peptides it reviewed — against the recommendation of the FDA's own career scientists, who advised against all seven. Many sites have reported this as "the FDA approved these peptides." That is false. Category 2 removal means only that a substance is no longer flagged as a significant safety risk, and an advisory vote is advisory. Lawful compounding requires notice-and-comment rulemaking, and the comparable prior process took the FDA over two years. Until a final rule issues, these substances satisfy none of the three conditions in section 503A and are not lawfully compoundable. Sermorelin is the exception, only because it was once a component of an approved drug.

Purity. Independent testing by BioTools found problems in nearly 30% of peptide vials tested: mislabelled, underdosed, overdosed, or contaminated with bacteria or toxins — their co-founder has described finding live or dead bacteria in a lot of vials, and noted this could lead to hospitalisations. Every dosing discussion above is partly hypothetical for anyone buying without a third-party certificate of analysis.

Cost. A year of a CJC-1295/ipamorelin protocol through a telehealth clinic runs $3,000 to $7,000, for a combination with no human body-composition trial. A year of gym membership and enough protein runs a few hundred dollars, for the intervention with the randomised trial and the bone density number.

How to decide

If muscle and bone are what you're worried about, this is the order we would put things in — and we are a site that earns money from peptide vendors.

1. Get a baseline. A DXA scan tells you whether this is normal density, osteopenia or osteoporosis, and that changes everything downstream. Ask, particularly within a couple of years either side of your final period.
2. Load the skeleton. Heavy progressive resistance training is the only intervention here with a randomised trial showing bone density moving the right way in women. If it is new to you, or you already have low bone mass, start supervised.
3. Eat enough protein. 1.0–1.2 g/kg as a floor, more if you train. Little on its own, a great deal alongside training.
4. Have the hormone therapy conversation properly — including route, which decides whether estrogen and any GH-directed peptide work with or against each other.
5. Consider the cheap adjuncts. Creatine and collagen peptides both have year-long randomised trials in postmenopausal women, both with modest effects and real limitations, both costing less than a month of any injectable here.
6. Only then look at peptides, eyes open — understanding that for muscle and bone you are paying for a mechanism, not a result, and the result has never been measured in anyone like you.

The same logic runs through our guides on energy and brain fog, sleep and libido and mood.

Questions people actually ask