What peptide kills fat cells?

The direct answer to the question is that no peptide has been shown to kill fat cells in humans. Part of the confusion is that "kills fat cells" conflates two very different biological events:

• Lipolysis — the fat cell releases its stored triglycerides and shrinks. The cell survives and can refill. This is what happens with diet, exercise, and every weight loss medicine with human evidence.
• Adipocyte death — the cell itself is destroyed, by programmed cell death (apoptosis) or by physical rupture of its membrane (lysis/necrosis). This is rare in therapeutics, and deliberately so: dying fat cells trigger inflammation, and nothing about an injected compound confines it to fat cells alone.

Compounds marketed as "lipolytic peptides" are, at best, claiming the first mechanism — and even that claim usually rests on animal data.

There is exactly one FDA-approved injectable that genuinely destroys fat cells: deoxycholic acid (Kybella). Its label describes it as "a cytolytic drug, which when injected into tissue physically destroys the cell membrane causing lysis" (Kybella prescribing information). Three details are instructive:

• It is a bile acid — a detergent-like molecule — not a peptide.
• It is approved only for submental fat (under the chin), and the label states that use elsewhere "has not been established and is not recommended."
• Its warnings show why "killing cells" is not gentle: reported harms include marginal mandibular nerve injury (asymmetric smile) and difficulty swallowing, plus near-universal injection-site swelling and bruising.

Unregulated "fat-dissolving" cosmetic products such as Lemon Bottle borrow this concept without the supporting trials, label, or quality control.

The compounds most often named as fat-cell killers have a much more modest literature:

• AOD-9604 (and its parent, HGH fragment 176-191) is a growth-hormone fragment designed to mimic hGH's fat-mobilizing effect. In obese mice, 14 days of treatment reduced body weight and fat via the beta-3 adrenergic receptor — a lipolysis pathway, not cell killing (Endocrinology 2001). Its human obesity program in the early 2000s never yielded published efficacy results or approval, and the compound is on WADA's prohibited list (Drug Test Anal 2015).
• 5-Amino-1MQ — a small molecule, not a peptide — inhibits the enzyme NNMT. In diet-induced obese mice it reduced fat mass and shrank adipocytes; the cells got smaller, they were not killed (Biochem Pharmacol 2018). No human trials have been published.

Meanwhile, the peptide drugs that actually reduce fat mass in humans do it without touching fat cells directly: in STEP 1 (n=1,961), semaglutide produced a mean 14.9% weight loss versus 2.4% for placebo by reducing appetite and food intake (NEJM 2021). The fat cells empty because the body runs an energy deficit — no cell death required.

The framing matters because it sets expectations the biology cannot meet. It suggests permanence ("dead cells can't come back") when the compounds being sold have not demonstrated fat-cell death in humans at all. It implies localized, targeted action, when injected compounds circulate systemically. And it makes unregulated vials sound more powerful than medicines that went through trials of thousands of participants — an inversion of the actual evidence.

A reliable rule of thumb: if a product's mechanism claim (cell death) is more dramatic than anything in its published literature (mouse lipolysis, or nothing), the claim is marketing.

Related reading: How do weight loss peptides work? and Do peptides really work for weight loss?