THE RESEARCH
Semaglutide research: the mechanism and the pivotal-trial record
From a 31-amino-acid peptide to a 17,604-person outcome trial — the evidence, organized by what each study measured.
Before the details
Semaglutide research spans the whole arc from a molecule on a bench to outcome trials with tens of thousands of people. The short version: it copies the gut hormone GLP-1, works mostly by acting on appetite circuits in the brain, and has been tested in three giant trial programs — SUSTAIN (diabetes by injection), PIONEER (diabetes by pill), and STEP (weight management). The weight numbers are large (about 15% in STEP 1 [2]). The outcome numbers are what set it apart: a 20% cut in heart events in SELECT [3], a 24% cut in kidney events in FLOW [5], and, in 2025, the first histology win in fatty-liver disease (MASH) [8].
Below, each major finding gets its own heading. Technical terms are glossed in plain words the first time they appear, and every number is tied to the study that produced it.
What is semaglutide
Semaglutide is a 31-amino-acid acylated analogue of human GLP-1 — meaning a lab-built copy of the natural gut hormone, modified to last far longer in the body [27]. It shares about 94% of its sequence with native GLP-1. Two changes do the heavy lifting: replacing the amino acid at position 8 with alpha-aminoisobutyric acid (Aib) blocks DPP-4, the enzyme that normally degrades GLP-1 within minutes; and attaching a C18 fatty di-acid chain lets the peptide bind tightly but reversibly to albumin (the most abundant blood protein), which slows its clearance [27].
The net effect of that engineering is an elimination half-life of about one week, the structural basis for once-weekly dosing [27]. It is supplied as a once-weekly subcutaneous injection and as a once-daily oral tablet co-formulated with the absorption enhancer SNAC [17].
How does semaglutide work
Semaglutide is a long-acting agonist of the GLP-1 receptor — it switches on the same receptor the natural hormone does, but for about a week instead of two minutes. Activating that receptor potentiates glucose-dependent insulin release from the pancreas, suppresses inappropriate glucagon (the hormone that raises blood sugar), and slows gastric emptying.
The weight-lowering effect, though, is mostly in the brain. Preclinical work in rodents showed semaglutide lowers body weight through distributed central nervous system pathways: it directly accessed the brainstem, area postrema (a region implicated in nausea and meal termination), hypothalamic arcuate nucleus and parabrachial nucleus, reduced food intake and modified food preference — without lowering energy expenditure [6]. In plain terms, it turns down hunger and the drive to eat rather than speeding up the body's calorie burn. The arcuate nucleus contains the satiety neurons (POMC/CART) it activates and the hunger neurons (NPY/AgRP) it inhibits.
Semaglutide peptide: the weight-management trials
As a semaglutide peptide therapy for weight, the pivotal result is STEP 1. In 1,961 adults with overweight or obesity and without diabetes, once-weekly subcutaneous semaglutide 2.4 mg produced a mean body-weight change of -14.9% from baseline to week 68, versus -2.4% with placebo — a treatment difference of about 12.4 percentage points [2]. The companion STEP 2 report extended the program into adults who also had type 2 diabetes [12].
It also improved the wider cardiometabolic picture: a STEP 1 and 4 exploratory analysis found improvements across waist circumference, blood pressure, lipids, glycemia and inflammatory markers versus placebo [15]. Durability and discontinuation are covered on the Semaglutide effects page — the short version is that the loss is sustained while the drug continues, and substantially regained when it stops [23][24].
Cardiovascular and kidney outcomes
This is the strongest part of semaglutide's evidence base, and the part tirzepatide cannot yet match with published hard-endpoint trials.
In SUSTAIN-6 (n=3,297 with type 2 diabetes at high cardiovascular risk), once-weekly semaglutide reduced the composite of cardiovascular death, nonfatal myocardial infarction or nonfatal stroke (HR 0.74; 95% CI 0.58-0.95) [4]. In SELECT (n=17,604 with established cardiovascular disease and overweight/obesity but no diabetes), semaglutide 2.4 mg reduced the same composite by 20% versus placebo (HR 0.80; 95% CI 0.72-0.90; P<0.001) [3] — the trial that established a cardiovascular indication independent of diabetes.
In FLOW (n=3,533 with type 2 diabetes and chronic kidney disease), once-weekly 1.0 mg reduced major kidney-disease events — kidney failure, a ≥50% decline in kidney function, or kidney/cardiovascular death — by 24% (HR 0.76; 95% CI 0.66-0.88) [5]. A meta-analysis of the oral formulation reported cardiovascular outcomes consistent with the GLP-1 receptor agonist class [11].
Liver disease (MASH): the 2025 readout
The newest indication came from the ESSENCE phase-3 trial (n=1,197; interim analysis n=800) in biopsy-confirmed metabolic dysfunction-associated steatohepatitis (MASH) — a progressive fatty-liver disease with inflammation and scarring. Once-weekly semaglutide 2.4 mg achieved resolution of steatohepatitis without worsening fibrosis in 62.9% of patients versus 34.3% with placebo, and fibrosis improvement without worsening steatohepatitis in 36.8% versus 22.4% [8]. Both differences were significant at P<0.001. This extended semaglutide's record from metabolic risk factors to a documented histologic (tissue-level) benefit in the liver.
Compounded semaglutide
Compounded semaglutide is semaglutide prepared by a compounding pharmacy rather than supplied as the approved manufactured product. During the federally declared shortage from roughly 2022 to early 2025, compounding pharmacies were permitted to produce it. Independent reporting and regulatory documentation during that period described dosing errors, adverse events requiring hospitalization, and products with unverified or non-pharmaceutical ingredients. After the shortage was declared resolved in 2025, those compounding pathways were curtailed, leaving ongoing regulatory uncertainty.
The key distinction for any reader: the entire trial record summarized on this site — STEP, SUSTAIN, SELECT, FLOW, ESSENCE — was generated with the approved manufactured product. Compounded or non-pharmaceutical sources fall outside that approved-product evidence base and carry documented quality and safety concerns. This digest summarizes the published research and takes no position on sourcing; it sells nothing and recommends nothing.