Semaglutide’s receptor chemistry has been mapped down to the amino acid. What has not been settled is where, in the body, the drug decides that you are full. The gut story, that a GLP-1 mimetic slows your stomach and that is why you eat less, is the version that fits in one sentence and gets repeated everywhere. Animal data point somewhere else: toward specific neurons in the hypothalamus and hindbrain, with the vagus nerve looking more like a bystander than a courier[1].
That distinction is not a footnote for pharmacologists. If the signal is decided in the brain rather than relayed from the gut, then every company trying to build the next GLP-1 drug without semaglutide’s gastrointestinal baggage is solving the wrong half of the problem.
What the STEP 1 trial measured
Before the mechanism, the outcome. In STEP 1, Wilding and colleagues randomly assigned 1,961 adults with obesity or overweight, without diabetes, to once-weekly subcutaneous semaglutide 2.4mg or placebo, alongside a lifestyle intervention, for 68 weeks[2].
14.9% average body weight lost at 68 weeks on semaglutide 2.4mg, vs 2.4% on placebo PMID 33567185
That design is randomized and controlled, the strongest structure available for this question. Participants did not choose to take the drug, and the trial did not simply track people who happened to already be using semaglutide. Randomization is what lets you attribute the weight loss to the drug itself, rather than to the kind of person who seeks out a new obesity treatment, or to the lifestyle counseling both groups received.
Interventional design has a blind spot, though: it tells you almost nothing about what happens once the intervention stops. STEP 1’s own extension answers that question, and not kindly. In the year after the 68-week treatment period ended, participants who had been on semaglutide regained an average of 11.6 percentage points of the weight they had lost, versus 1.9 points among former placebo recipients[3]. The appetite effect fades once the drug does, closer to a lease than a cure.
Bottom line: the STEP 1 data support a causal, clinically large effect while the drug is being taken. They do not support a durable one once it is stopped.
The chemistry that buys semaglutide a week instead of minutes
Semaglutide is a modified analog of GLP-1, carrying two amino acid substitutions, Aib8 and Arg34, that block the enzymatic degradation the native hormone is normally subject to, plus a fatty acid chain attached at lysine 26 that binds circulating albumin, described by Lau and colleagues in the paper reporting the drug’s discovery[4]. That albumin tether is the whole trick: bound to albumin, semaglutide is too large to be cleared quickly by the kidneys and too well hidden to be cut apart by the enzyme that dismantles native GLP-1. Human pharmacokinetic studies report a plasma half-life for semaglutide of approximately one week, far longer than native GLP-1, which is rapidly broken down by the enzyme DPP-4, and this extended half-life is what makes once-weekly dosing possible[5].
Once it is circulating, semaglutide does what GLP-1 does, just for far longer. On pancreatic beta cells, receptor activation increases insulin secretion in a glucose-dependent way, meaning it does not push insulin release when glucose is already low, and it reduces circulating glucagon[6]. That is the diabetes mechanism, and it existed as a reason to develop the drug well before semaglutide became a weight-loss headline.
Where the appetite signal starts
Here is the part that gets flattened in most explainers: semaglutide slows gastric emptying and dampens appetite through GLP-1 receptors that sit in both the gastrointestinal tract and the central nervous system[7]. That sentence is true and still unhelpfully vague, because “the central nervous system” is doing a lot of work in it.
GLP-1 receptors are not spread evenly through the brain. In mice, they cluster in specific hypothalamic and hindbrain sites tied to appetite control: the arcuate nucleus, the area postrema, the paraventricular nucleus, and the ventromedial hypothalamus[8].
The receptor map argues against a gut-only story.
This is where one specific rat experiment matters more than any class-wide review. Secher and colleagues, testing liraglutide (an earlier and shorter-acting GLP-1 receptor agonist), found that weight loss in rats did not require GLP-1 receptors in the vagus nerve, the area postrema, or the paraventricular nucleus at all[1]. The effect instead depended on the drug binding directly to POMC and CART neurons inside the arcuate nucleus, and that binding disappeared in mice engineered to lack the GLP-1 receptor entirely. The vagus nerve, so central to the “your gut talks to your brain” version of this story, was not needed.
No one has run the human equivalent of that experiment, for the obvious reason that you cannot lesion a hindbrain nucleus in a person to see what happens to their appetite. The dispute over how much of a GLP-1 receptor agonist’s effect is peripheral versus central is, right now, confined to rodent circuitry[1]. Every popular account that reduces the mechanism to “it slows your stomach” is quietly resolving a question that has not been resolved.
How the trials dosed it
The published protocols are research parameters. They describe what investigators tested inside a monitored trial, and copying them at home is a different exercise entirely. STEP 1 built in a 16-week dose-escalation period before participants reached the 2.4mg weekly maintenance dose[3], a ramp widely assumed to blunt the gastrointestinal side effects that show up when a GLP-1 receptor agonist is dosed at full strength from the start.
Semaglutide’s other delivery problem is oral bioavailability, and it is a much older problem than semaglutide itself. Insulin has faced the same wall for a century: a peptide swallowed whole gets shredded by stomach acid and digestive enzymes before it reaches the bloodstream in any useful amount. Semaglutide’s workaround is a co-formulation with the absorption enhancer SNAC, and even with that enhancer, the oral formulation tested in the placebo-controlled OASIS 1 trial needed a 50mg daily dose[9] to approximate what 2.4mg does by injection. That gap between an oral dose and an injectable one is the bioavailability tax any oral peptide pays, semaglutide included.
The side effects the trials recorded
In a phase 2 dose-ranging trial that put semaglutide, liraglutide, and placebo through the same protocol, O’Neil and colleagues found semaglutide doses of 0.2mg or higher produced mean weight loss of 11.2% to 13.8% at 52 weeks, against 7.8% for liraglutide 3.0mg[10]. The dose-response data from that trial is worth seeing in full:
Weight change by treatment arm at 52 weeks (phase 2 dose-ranging trial)
Placebo: -2.3
Semaglutide 0.05mg: -6
Semaglutide 0.1mg: -8.6
Semaglutide 0.2mg: -11.6
Semaglutide 0.3mg: -11.2
Semaglutide 0.4mg: -13.8
Liraglutide 3.0mg: -7.8
Placebo, ascending daily semaglutide doses, and liraglutide 3.0mg, all from the same trial.
Gastrointestinal complaints dominate the safety record elsewhere in the literature, and they are not rare. A meta-analysis of 13 randomized trials in adults with obesity found semaglutide increased the risk of gastrointestinal adverse events, mainly nausea, vomiting, and diarrhea, by a relative risk of 1.68 compared with placebo[11].
The gallbladder signal is smaller in absolute terms but harder to dismiss:
2.6x increased risk of gallbladder disorders (cholelithiasis) vs placebo PMID 40189856
Gastrointestinal events: mostly transient and concentrated in the dose-escalation window, but the leading reason people discontinue the drug.
Gallbladder disease: a real, statistically significant elevation in risk that is easy to underestimate.
Everything else: the trials behind these numbers monitored participants on a fixed visit schedule. The safety profile describes supervised use. It says nothing about someone titrating alone off a compounding-pharmacy vial.
The leap people skip: every adverse-event figure above comes from a monitored trial population. Calling the side-effect profile “mild and manageable” for unsupervised use is exactly the extrapolation the data does not support.
Semaglutide against liraglutide and tirzepatide
Semaglutide did not arrive in a vacuum. It replaced liraglutide on the strength of the numbers above, and it is now being outcompeted in turn.
A meta-analysis of direct comparative studies between semaglutide and tirzepatide, including two randomized trials among seven studies overall, found that tirzepatide, a dual GLP-1 and GIP receptor agonist, produced significantly greater weight loss than semaglutide[12]. The authors are careful to note that long-term head-to-head evidence is still thin, which is a more honest claim than a flat “tirzepatide wins,” but the direction of the result has held up every time someone has run the comparison.
If GIP co-agonism is genuinely adding an effect on top of GLP-1 receptor activation, and tirzepatide’s edge suggests something is, the interesting question stops being how semaglutide works and becomes how much of the weight-loss ceiling GLP-1 alone was ever going to reach. For the fuller trial-by-trial comparison, see how semaglutide, tirzepatide, and retatrutide stack up against each other.
What this settles, and what it does not
The efficacy question is closed by trial standards. Semaglutide produces large, dose-dependent, randomized-trial-confirmed weight loss, through a well-mapped set of GLP-1 receptor actions on the pancreas, the gut, and a specific handful of brain nuclei[8].
The mechanism question is not closed. Nobody has directly tested, in a living person, how much of the appetite effect would survive losing vagal signaling versus losing central GLP-1 receptor activity. Until an experiment like that exists in some form, “semaglutide works by slowing your stomach” is a simplification that happens to be easy to say, and the rodent data available right now only supports it as part of the picture.
That gap is worth tracking, not because it changes anything for someone deciding whether to take the drug today, but because the next generation of anti-obesity peptides is being engineered around assumptions about where this signal originates. If the arcuate nucleus turns out to matter more than the gut, drugs designed mainly to avoid crossing into the brain are optimizing for the wrong target. For the broader trial record behind semaglutide, the five trials that define its real benefits is a useful next stop, and anyone considering a compounded version should read what changes when the vial doesn’t match the trial before assuming the dosing above transfers directly.
This article is for research and informational purposes only and is not intended to diagnose, treat, cure, or prevent any disease. The peptides discussed here are sold for research use only and are not for human consumption. Nothing in this article constitutes medical advice. Consult a qualified clinician before making changes to a health, training, or supplementation protocol.
References
The arcuate nucleus mediates GLP-1 receptor agonist liraglutide-dependent weight loss.. The Journal of clinical investigation, 2014.
Once-Weekly Semaglutide in Adults with Overweight or Obesity.. The New England journal of medicine, 2021.
Weight regain and cardiometabolic effects after withdrawal of semaglutide: The STEP 1 trial extension.. Diabetes, obesity & metabolism, 2022.
Discovery of the Once-Weekly Glucagon-Like Peptide-1 (GLP-1) Analogue Semaglutide.. Journal of medicinal chemistry, 2015.
Safety and Pharmacokinetics of Single and Multiple Ascending Doses of the Novel Oral Human GLP-1 Analogue, Oral Semaglutide, in Healthy Subjects and Subjects with Type 2 Diabetes.. Clinical pharmacokinetics, 2019.
Anti-diabetic actions of glucagon-like peptide-1 on pancreatic beta-cells.. Metabolism: clinical and experimental, 2014.
User's guide to mechanism of action and clinical use of GLP-1 receptor agonists.. Postgraduate medicine, 2015.
Identification and characterization of GLP-1 receptor-expressing cells using a new transgenic mouse model.. Diabetes, 2014.
Oral semaglutide 50 mg taken once per day in adults with overweight or obesity (OASIS 1): a randomised, double-blind, placebo-controlled, phase 3 trial.. Lancet (London, England), 2023.
Efficacy and safety of semaglutide compared with liraglutide and placebo for weight loss in patients with obesity: a randomised, double-blind, placebo and active controlled, dose-ranging, phase 2 trial.. Lancet (London, England), 2018.
Gastrointestinal safety of semaglutide and tirzepatide vs. placebo in obese individuals without diabetes: a systematic review and meta analysis.. Annals of Saudi medicine, 2025.
