Amylin Analogs and GLP-1 Microdosing: What Cagrilintide, CagriSema and Amycretin Add to the Picture

Every time a meal is eaten, the pancreas releases two hormones from the same cells at the same moment. One is insulin. The other is amylin, and for most of the GLP-1 era it has been the hormone nobody outside a pharmacology department talked about.

That changed with a set of trial results between 2021 and 2025. Cagrilintide, a once weekly amylin analog, produced about 10.8 percent weight loss on its own over 26 weeks. Paired with semaglutide 2.4 mg in a fixed combination called CagriSema, it produced about 20.4 percent over 68 weeks in a phase 3 trial of 3,417 people, against 14.9 percent for semaglutide alone. A single molecule that hits both receptors, amycretin, reached about 24.3 percent in an early trial. Three other companies now have amylin analogs in mid stage development.

The reason this belongs on a site about low dose GLP-1 protocols is not that anyone should be combining peptides at home. It is that the amylin trials are the clearest demonstration yet of two principles that underpin low dose thinking: that dose response curves for these hormones are front loaded, and that adding a second pathway at a modest dose can do more than pushing the first one to its ceiling. The article below sets out what amylin is, what the trials measured, and where the evidence stops.

What is amylin and how is it different from GLP-1?

Same meal, different gland, different part of the brain

Amylin is a 37 amino acid peptide hormone stored in the same secretory granules as insulin inside pancreatic beta cells and released with it, at roughly one molecule of amylin for every hundred of insulin, whenever glucose rises after a meal. It was identified in 1987 as the main component of the amyloid deposits found in the pancreas of people with type 2 diabetes, which is where the name comes from, and its role as a satiation hormone was worked out over the following decade. GLP-1, by contrast, is made in the L cells of the gut lining and released in response to food arriving in the intestine. The two hormones rise together at every meal but come from different organs.

Amylin does three things that overlap with GLP-1 and one that does not. It slows gastric emptying, it suppresses the glucagon surge that follows a meal, and it reduces food intake by signalling that a meal is complete. The distinct part is where it acts. Amylin's primary target is the area postrema, a small region on the floor of the fourth ventricle in the brainstem that sits outside the blood brain barrier and samples circulating hormones directly. From there the signal is relayed to the lateral parabrachial nucleus and onward to the hypothalamus. GLP-1 receptor agonists reach the area postrema too, but their strongest appetite effects run through the arcuate nucleus of the hypothalamus and the nucleus of the solitary tract, and through reward circuits in the midbrain.

The amylin receptor is also unusual. It is not a single protein but a calcitonin receptor paired with one of three receptor activity modifying proteins, producing three receptor subtypes named AMY1, AMY2 and AMY3. Because the calcitonin receptor is part of the complex, some amylin drugs activate the bare calcitonin receptor as well, and the drugs in development differ on exactly this point. Cagrilintide is a dual amylin and calcitonin receptor agonist. Petrelintide and eloralintide were designed to be more selective for the amylin receptor complex, on the theory that the calcitonin component contributes to nausea without adding much to appetite.

Why did it take twenty years to get from pramlintide to cagrilintide?

The first amylin drug worked, but three injections a day was too many

The first amylin analog, pramlintide, was approved in 2005 for people with type 1 or type 2 diabetes who use mealtime insulin, and it is still the only amylin drug on the market. Human amylin cannot be used as a drug because it clumps into the same amyloid fibrils that gave it its name, so pramlintide swaps three amino acids for proline to stop the aggregation. It works: it flattens post-meal glucose, lowers glucagon, and in 16 week obesity trials at 240 to 360 micrograms three times daily it produced about 3.7 percent weight loss against placebo. Paired with the leptin analog metreleptin in a 2008 proof of concept study, it produced about 12.7 percent at 20 weeks, an early signal that amylin combines well with other pathways.

The problem was practical. Pramlintide has a half-life of about 48 minutes and has to be injected before every major meal, three times a day, from a separate pen to the insulin it is taken alongside. Adherence was poor and the drug never sold in volume. Building a once weekly version required the same protein engineering that turned liraglutide into semaglutide: attaching a fatty acid chain that binds albumin in the blood and protects the peptide from clearance, while modifying the sequence so it stays soluble and stable in a pen.

Cagrilintide is the result. It is a 37 amino acid analog with a C20 diacid lipid attached, giving it a half-life of around a week. The phase 2 trial published in 2021 was the first time a long acting amylin analog had been tested in people with obesity at scale, and the fact that it matched liraglutide 3.0 mg at 2.4 mg and beat it at 4.5 mg is what moved the whole field. Within four years Zealand Pharma, Eli Lilly and AstraZeneca each had an amylin analog in clinical trials.

Why does adding amylin beat raising the GLP-1 dose?

A second pathway has no ceiling in common with the first

Adding an amylin analog to semaglutide produces more weight loss than any increase in semaglutide dose can, because the two hormones act on different neurons and the GLP-1 pathway has already flattened by 2.4 mg. Every GLP-1 receptor agonist dose response curve has the same shape: steep at the bottom and flat at the top. Semaglutide 1.0 mg produces roughly two thirds of the weight loss of 2.4 mg, and doses above 2.4 mg add little beyond more nausea. Pushing a single receptor harder runs into receptor saturation, tolerance, and the limits of what one circuit can do. A second receptor on a second population of neurons starts its own curve from zero.

The phase 1b combination trial published in 2021 makes this unusually visible. Ninety-six adults with overweight or obesity received semaglutide escalated to 2.4 mg with either placebo or one of five cagrilintide doses added, from 0.16 mg to 2.4 mg, for 20 weeks. Semaglutide with placebo produced about 9.8 percent weight loss. Semaglutide with the smallest cagrilintide dose, 0.16 mg, produced about 15.7 percent. The largest dose, 2.4 mg, produced about 17.1 percent. Almost the whole benefit of the second hormone arrived with the first and smallest step, and a fifteen-fold increase in the amylin dose bought only another 1.4 percentage points.

Animal work suggests why. Amylin appears to restore sensitivity to leptin, the fat-derived hormone that the brain stops responding to in obesity, and rats given amylin plus leptin lose more fat than either alone. Amylin also reduces the reward value of food through a different route from GLP-1's dopamine effects. The two signals converge on the same hypothalamic output but arrive from different directions, which is the definition of a complementary rather than a redundant pairing.

Which amylin drugs are in development and what have they shown?

Five agents, four companies, one approved and four in trials

Five amylin-based agents have produced human weight loss data as of 2026, and they fall into three groups: amylin analogs given alone, amylin analogs paired with a GLP-1 in one injection, and single molecules that activate both receptors. The trial durations and populations differ enough that the numbers below are a map of the field rather than a ranking.

Do amylin analogs cause less nausea than GLP-1 medications?

Less on their own, about the same in combination

Amylin analogs given alone cause fewer gastrointestinal side effects than GLP-1 receptor agonists at similar weight loss, and the combination adds little on top of what semaglutide already causes. In REDEFINE 1, roughly 63 percent of participants on cagrilintide alone reported a gastrointestinal adverse event over 68 weeks, against about 76 percent on semaglutide and about 80 percent on CagriSema, with placebo at around 40 percent. Nausea specifically ran at roughly 30 percent on cagrilintide against the mid 40s on either semaglutide arm. Those are approximate figures, and the pattern is that the amylin arm sat halfway between placebo and the GLP-1.

The nausea that amylin does cause comes partly from the same place its appetite effect does. The area postrema is also the chemoreceptor trigger zone, the brainstem region that initiates vomiting in response to circulating toxins, and any drug that stimulates it strongly will produce some nausea. The selective agonists were designed on the hypothesis that the calcitonin receptor component of cagrilintide's activity adds to the nausea without adding to satiation. The early petrelintide and eloralintide data are consistent with that, though the trials are small and short and the comparison has not been made directly.

One question that matters for the long term is lean mass. Weight loss on any agent includes some muscle, and the proportion on semaglutide is around a quarter to a third of total loss, as covered elsewhere on this site. Zealand Pharma has argued that amylin analogs produce a higher proportion of fat loss, based on animal data and body composition sub-studies, and the REDEFINE 1 DEXA substudy is the first large human dataset on the question. The published proportion of lean mass lost on CagriSema was similar to semaglutide rather than better, which means the lean mass advantage remains a hypothesis rather than a result.

What do the amylin trials mean for low dose GLP-1 thinking?

Two principles, demonstrated with someone else's molecule

The amylin trials support two ideas that low dose GLP-1 protocols rest on, without saying anything about any particular protocol. The first is that the dose response for these hormones is front loaded. Cagrilintide at 0.3 mg delivered 56 percent of the effect of the 4.5 mg dose. Semaglutide at 0.5 mg delivers a substantial fraction of the effect at 2.4 mg. Tirzepatide at 5 mg delivers roughly three quarters of the effect at 15 mg. The shape is the same across three different molecules and two different receptor systems, and it means that the first milligram of any of these drugs is worth far more than the last.

The second is that a modest amount of a second mechanism can add more than a large increase in the first. The combination trial's 0.16 mg arm is the cleanest example in the literature: a dose one fifteenth of the eventual product dose, added to a GLP-1 already at its ceiling, produced most of the additional weight loss that the full combination did. The same logic underlies tirzepatide, where GIP receptor activity added to GLP-1 outperforms GLP-1 alone at any dose, and it is the reason the industry is now building triple agonists rather than stronger single agonists.

What the trials do not support is any version of assembling that combination outside a trial. Cagrilintide has been studied at one fixed dose, in one fixed pairing, under escalation schedules designed by the sponsor and monitored by trial clinicians. It is not an approved medicine, it is not available through any pharmacy, and the products sold online under its name have no verified identity, purity, or concentration. The relevance of the amylin data to a low dose protocol is that they confirm the shape of the curves. They do not add an ingredient.

What is worth knowing and raising with a prescriber?

Five points, none of which are a protocol

Everything below is context for a conversation rather than an instruction. Nothing here sets a dose, adds a medication, or replaces an assessment.

What is genuinely unresolved here?

Three gaps worth naming

The low dose combination data come from one trial with sixteen people per arm. The finding that 0.16 mg of cagrilintide added most of what 2.4 mg added is striking, and it is also exactly the kind of result that a small trial produces by chance. The phase 3 programme tested only the 2.4 mg dose of each component, so nobody has measured whether a lower cagrilintide dose in combination holds up over a year in a large population. The 0.16 mg result is a hypothesis about curve shape, not a dosing recommendation.

The lean mass question is open. Amylin analogs have been promoted as producing higher quality weight loss, meaning a larger share from fat, on the basis of animal studies and small body composition datasets. The one large human measurement, the REDEFINE 1 DEXA substudy, did not show a clear advantage over semaglutide. Whether the selective agonists do better, and whether the difference is large enough to matter, will not be known until their phase 3 body composition data are published.

Long term safety of chronic amylin receptor stimulation has not been established beyond 68 weeks. Pramlintide has two decades of use in diabetes without a signal beyond nausea and insulin-related hypoglycaemia, but at a fraction of the receptor exposure a weekly long acting analog produces. The calcitonin receptor component of cagrilintide's activity, its effect on bone over years, and whether any of the amyloid biology that gave amylin its name is relevant to a stabilised analog are all questions the current trials were not designed to answer.

Frequently Asked Questions