Are Peptides Safe Long Term? What Is Known, What Is Not, and the Real Risks

SELECT, in the New England Journal of Medicine in 2023, followed people with obesity and established cardiovascular disease for years on semaglutide; STEP 5 in Nature Medicine in 2022 ran two years. For the majority of compounded peptides, no trial of this kind exists at any duration.

Once a product is approved, clinicians and patients report problems through a national system and the label is revised as signals accumulate. The gallbladder and pancreatitis language on the approved semaglutide and tirzepatide labels is that process working. Compounded preparations sit outside the same machinery.

An overview in the Journal of Clinical Endocrinology and Metabolism in 2022 pooled long-term safety follow-up on more than fifteen thousand adults treated with recombinant growth hormone for growth hormone deficiency. That dataset is real, and it describes recombinant growth hormone in diagnosed deficiency rather than compounded secretagogues in healthy adults.

A peptide with none of the three has mechanism, animal work and prescriber observation. That is how a great deal of medicine is practised, and it is not nothing. It is also not a safety dataset, and calling it one is the specific dishonesty this page exists to avoid.

This class sits in a different category from everything else here: STEP 1, STEP 5 at two years, SUSTAIN-6, SELECT and ESSENCE for semaglutide, and SURMOUNT-1, SURPASS-2, SURMOUNT-OSA and SURPASS-CVOT for tirzepatide — thousands of randomized participants followed for years, published between 2016 and 2025. The approved labels for both actives carry a boxed warning about thyroid C-cell tumors observed in rodents, with both stating that the human relevance of that finding has not been determined, and both are contraindicated in people with a personal or family history of medullary thyroid carcinoma or with multiple endocrine neoplasia syndrome type 2. Labeled warnings cover acute pancreatitis, acute gallbladder disease, acute kidney injury from volume depletion, hypoglycemia alongside insulin or an insulin secretagogue, diabetic retinopathy complications, heart-rate increase, and monitoring for depression or suicidal thoughts; the tirzepatide label adds severe gastrointestinal reactions, a note that it is not recommended in severe gastroparesis, and pulmonary aspiration during general anesthesia or deep sedation.

Two findings matter more to most people than any of that. On stopping, the STEP 1 trial extension in Diabetes, Obesity and Metabolism in 2022 described weight and cardiometabolic measures moving back toward baseline in the following year. On body composition, a systematic review in Annals of Internal Medicine in 2026 examined how incretin-based and non-pharmacologic weight loss affect lean mass. Semaglutide and tirzepatide as compounded preparations share the active ingredient but not the approval.

CJC-1295 was shown to raise growth hormone and IGF-1 in healthy adults in the Journal of Clinical Endocrinology and Metabolism in 2006, ipamorelin was characterised as the first selective growth hormone secretagogue in the European Journal of Endocrinology in 1998, and sermorelin has human trial work including long-term administration in age-advanced adults in 1997 and a controlled cognition trial in Archives of Neurology in 2012. Hexarelin has the longest human endocrine record of the group, including a 1998 study of growth hormone status during long-term hexarelin therapy describing the GH response attenuating under sustained exposure, and a review in Sexual Medicine Reviews in 2018 is the best single summary of the class. What none of it provides is a multi-year outcome study of a compounded secretagogue in a healthy adult.

Tesamorelin is the exception — the one GH-axis molecule with an FDA-approved finished product, trialled in the New England Journal of Medicine in 2007, with safety extension data in 2010 and a fatty liver trial in Lancet HIV in 2019. Its label states the class caution directly: tesamorelin raises serum IGF-1, "the effects of prolonged elevations in IGF-1 levels are unknown," and IGF-1 should be monitored with discontinuation considered where elevations persist. It also warns on fluid retention including edema, arthralgia and carpal tunnel syndrome, and on glucose intolerance, and is contraindicated in active malignancy, in disruption of the hypothalamic-pituitary axis, and in pregnancy. CJC-1295 / Ipamorelin, sermorelin, ipamorelin and tesamorelin are managed to that logic: IGF-1 inside a physiological range, fasting glucose watched, cancer history screened at intake.

IGF-1 LR3 does not stimulate your own axis — it supplies the growth factor directly, engineered to evade the binding proteins that normally restrain it, as described in the Journal of Molecular Endocrinology in 1992. A meta-regression in the Lancet in 2004 examined circulating IGF-1 in relation to cancer risk across populations; that is epidemiology of endogenous levels rather than a study of injected IGF-1 LR3, but it is why prescribers are more conservative here than anywhere else. A 2010 paper in Growth Hormone and IGF Research reported detecting a His-tagged Long-R3-IGF-I, a manufacturing-tag impurity, in a black-market product.

There are no randomised controlled trials of BPC-157 in humans for tendon, muscle or soft-tissue injury. The evidence is animal and cell work — Achilles tendon detachment in rats in the Journal of Orthopaedic Research in 2006, the gastrointestinal literature reviewed in Current Pharmaceutical Design in 2011. TB-500 is the one repair peptide with any published human safety study: a randomised, placebo-controlled single and multiple dose study of intravenous thymosin beta-4 in healthy volunteers, in the Annals of the New York Academy of Sciences in 2010 — a short phase 1 study. KPV has the thinnest file, with anti-inflammatory activity described in Gastroenterology in 2008. The theoretical long-term concern for BPC-157 and TB-500 is angiogenesis, since blood supply serves any tissue that needs it; nothing published reports this becoming a clinical problem, and nothing published could, because the studies that would detect it over years have not been run. That is why a cancer history is discussed at intake and why KPV, which works through melanocortin signalling instead, is sometimes the better choice. See can peptides cause cancer.

The long-term question for GHK-Cu is not the peptide but the metal it carries. Collagen stimulation in cultured fibroblasts was reported in FEBS Letters in 1988 and the gene-expression case is set out in the International Journal of Molecular Sciences in 2018, but most human GHK-Cu work is topical or in wound care. Copper is an essential trace element with a narrow useful range, and no study has characterised the cumulative burden of years of injected GHK-Cu — which is why liver function is checked at baseline, why copper-handling disorders are exclusions, and why total exposure is reviewed before a second copper-containing product is added. Is GHK-Cu safe covers this properly.

MOTS-c was characterised in Cell Metabolism in 2015 and as an exercise-responsive regulator in Nature Communications in 2021 — animal and cell work, with no long-term human administration data for MOTS-c. Epithalon has a Russian literature going back decades, including telomerase activity in human somatic cells in 2003 and a 2025 replication in Biogerontology; those human follow-up studies are small and not independently replicated, so it belongs in the preclinical tier. SS-31 is the exception: as elamipretide it has a randomised crossover trial in primary mitochondrial myopathy in 2020, the MMPOWER-3 trial in Neurology in 2023 — which did not meet its primary endpoint — and a Barth syndrome trial in Genetics in Medicine in 2021 with 168-week open-label extension results in 2024. An approved product now exists for Barth syndrome, its label listing injection-site reactions as the most common adverse reactions with a hypersensitivity warning.

Thymosin alpha-1 has the largest clinical file of the immune group, reviewed in the World Journal of Virology in 2020, and is an approved product in a number of countries outside the United States. Thymalin's literature is almost entirely Russian. The long-term consideration here is not toxicity — it is that chronic immune modulation is a different proposition in someone with an autoimmune condition or on an immunosuppressant, which makes Thymosin Alpha-1 a specialist conversation before it is a prescription.

These sit at the safest end of the page for an unglamorous reason: the body already handles them in quantity, with decades of clinical use behind the deficiency indications. Methylcobalamin is vitamin B12 — deficiency and repletion are covered in the New England Journal of Medicine in 2013 and in British Journal of Haematology guidelines in 2014, and methylcobalamin has the most established long-term profile of anything Pepti supplies. L-Carnitine has meta-analyses in Obesity Reviews in 2016 and Pharmacological Research in 2020, and one genuine long-term question that is metabolic rather than toxic: gut-microbiome conversion of carnitine to TMAO and its relationship to atherosclerosis, in Nature Medicine in 2013.

NAD+ has a large human literature on precursors — nicotinamide riboside in Nature Communications in 2018, an NMN dose-ranging trial in GeroScience in 2023 — plus a pilot on NAD+ metabolism during intravenous infusion in Frontiers in Aging Neuroscience in 2019. The honest gap is that years-long outcome data for injected NAD+ does not exist, and NAD+ metabolism intersects with proliferation and DNA-repair pathways, which is why a cancer history is discussed. Glutathione has randomized human work in the European Journal of Nutrition in 2015 and Parkinson's trials in Movement Disorders in 2009, and every cell makes it continuously — the strongest structural argument for its tolerability, and still not a long-term injectable outcome study.

PT-141 is bremelanotide, one of the few peptides here with a published long-term safety study by name, in Obstetrics and Gynecology in 2019. The approved label is specific: a transient rise in blood pressure and fall in heart rate after each dose, usually resolving within about twelve hours; focal hyperpigmentation reported by 1% of patients receiving up to eight doses per month, with higher risk in people with darker skin and with daily dosing; and nausea reported by 40% of patients. It is contraindicated in uncontrolled hypertension or known cardiovascular disease. PT-141 is used episodically rather than daily, which is itself why its long-term exposure question is smaller than most.

Oxytocin has a long obstetric history as an approved injectable and a large intranasal research literature, including a randomized trial of long-term intranasal oxytocin in Fertility and Sterility in 2015 — a literature also notable for methodological criticism, with two papers in Biological Psychiatry in 2016 arguing intranasal findings have been over-interpreted. Kisspeptin has randomized trials in hypoactive sexual desire disorder in JAMA Network Open in 2022 and 2023, and one directly relevant long-term finding: chronic administration of kisspeptin-54 caused tachyphylaxis in women with hypothalamic amenorrhea, in the Journal of Clinical Endocrinology and Metabolism in 2009.

Semax and Selank have a Russian clinical literature that is genuinely clinical and genuinely hard to audit from outside — selank compared head to head with phenazepam in Zhurnal Nevrologii i Psikhiatrii in 2014, semax with stroke trials in the same journal — without the independent replication that would let anyone quote a long-term profile. FOXO4-DRI is a senolytic whose founding paper in Cell in 2017 is a landmark, and whose honest long-term note comes from the senolytic field itself: a 2023 paper in Circulation reported that eliminating senescent cells promoted pulmonary hypertension development and progression in a model system.

Melanotan II is the one molecule here where published human case reports rather than theory define the caution — a review of the risks of unregulated alpha-MSH analogue use in the International Journal of Dermatology in 2017, and case reports in Dermatology in 2014 and the Australasian Journal of Dermatology in 2012 describing melanoma and melanoma in situ associated with use — so anyone considering Melanotan II needs a dermatological assessment first. AOD-9604 appears in a 2026 review of unapproved musculoskeletal peptides in Sports Medicine, and 5-Amino-1MQ rests on NNMT inhibition work in Biochemical Pharmacology in 2018 and Nature in 2014. Long-term human data for AOD-9604 and 5-Amino-1MQ alike is zero.

Very little. The list lives in federal regulation, and as it currently stands it contains six substances: Brilliant Blue G, cantharidin, diphenylcyclopropenone, N-acetyl-D-glucosamine, squaric acid dibutyl ester and thymol iodide — four of them topical-use-only, and not one of them a peptide. FDA's page describing the list and the interim policies is current as of 14 May 2026.

Category 2 is the interim-policy grouping for substances FDA has identified as potentially presenting significant safety risks. It currently holds, under the 503A interim policy, cesium chloride, domperidone, germanium sesquioxide, ibutamoren mesylate, kisspeptin-10 and quinacrine hydrochloride; ipamorelin acetate is in Category 2 under the 503B policy, which applies to outsourcing facilities rather than 503A prescription compounding. Kisspeptin-10 is the one that touches this catalog, and it is there because FDA has limited safety information for the proposed routes, not because a safety problem was observed.

A long list of peptides previously in Category 2 now sits in a separate table on the same FDA page, headed "bulk drug substances nominated but withdrawn": AOD-9604, BPC-157, cathelicidin LL-37, CJC-1295, dihexa acetate, emideltide (DSIP), epitalon, GHK-Cu for injectable routes, ipamorelin acetate, KPV, PEG-MGF, melanotan II, MOTS-c, selank acetate, semax, thymosin alpha-1 and the thymosin beta-4 fragment sold as TB-500.

They came off because the nominations were withdrawn by the nominators, not because FDA reviewed them again and cleared them. FDA's stated concerns are still published beside each one and read as a group they are consistent: potential immunogenicity for certain routes, complexity in characterising peptide-related impurities, and — repeatedly — no or limited safety information. That is a statement about missing data, not a finding of harm. Anyone quoting the old "Category 2" framing is out of date; anyone quoting the removal as an FDA endorsement is misrepresenting it.

FDA's Pharmacy Compounding Advisory Committee met on 23–24 July 2026 to consider seven bulk substances for inclusion on the 503A Bulks List: BPC-157 (nominated for ulcerative colitis), KPV (wound healing and inflammatory conditions), TB-500 (wound healing) and MOTS-c (obesity and osteoporosis) on the first day; emideltide (opioid withdrawal, chronic insomnia and narcolepsy), semax (cerebral ischemia, migraine and trigeminal neuralgia) and epitalon (insomnia) on the second.

A committee recommendation is not a determination. FDA issues its own decision afterwards, and none of these substances appears on the list in federal regulation as it currently stands. The accurate description today is: legal to prescribe and dispense under section 503A, not on the Bulks List, no longer in Category 2, awaiting FDA action.

Compounded medications are prepared by a licensed pharmacy pursuant to a prescription for an individual patient. They are not approved as finished drug products and never will be, because that is not what compounding is. Federal regulation states that representing a compounded drug made with a listed bulk substance as FDA approved would misbrand it, so anyone marketing a compounded peptide as FDA approved is making a claim the regulation forecloses. Are peptides FDA approved covers the distinction. Separately, if you compete under anti-doping rules: a 2026 WADA Prohibited List is in effect and a 2027 list takes effect on 1 January 2027. Many peptides are prohibited and the list changes annually, so check the current list for your sport and tell your physician you are tested.

The point of baseline labs is not the peptide — it is finding what symptoms hide. Thyroid disease, undiagnosed glucose problems, anemia, and liver or kidney impairment all present as fatigue, and fatigue is what brings many people to a peptide in the first place. At-home blood testing covers the markers used, and do you need bloodwork before peptides explains when a physician will insist.

The single most useful long-term number in the catalog, and the approved tesamorelin label says why: the effects of prolonged IGF-1 elevation are unknown, so the value is monitored and persistent elevation is a reason to reconsider. A prescriber managing sermorelin, CJC-1295, ipamorelin, tesamorelin or IGF-1 LR3 without ever looking at IGF-1 is not monitoring you.

Growth hormone signalling opposes insulin, and the approved tesamorelin label directs glucose evaluation before and during therapy; on the GLP-1 side the concern runs the other way, and the same two numbers answer both questions. Liver and kidney function matter to GHK-Cu because of the copper, to NAD+ and glutathione because of clearance, and to the GLP-1 class because volume-depleting gastrointestinal effects can affect renal function, which the approved semaglutide label names specifically.

The approved bremelanotide label directs consideration of cardiovascular risk before starting PT-141 and periodically during treatment, and states it is not recommended in people at high risk of cardiovascular disease. For melanotan II the most important item of monitoring is not a lab at all — it is a dermatological skin check.