Semaglutide is a once-weekly GLP-1 receptor agonist (Ozempic for T2D, Wegovy for obesity, Rybelsus as a daily oral tablet) with the deepest clinical data in the incretin class. SELECT followed 17,604 adults with obesity and established CVD for ~4 years on 2.4 mg and delivered a 20% reduction in major cardiac events vs placebo. FLOW showed kidney benefit at 1.0 mg in T2D + CKD; ESSENCE produced 62.9% MASH resolution at 2.4 mg vs 34.1% placebo at 72 weeks; STEP-2 directly compared 1.0 vs 2.4 mg. The ~4-year SELECT exposure gives a safety record with no peer in the class. What semaglutide does NOT have is data above 2.4 mg weekly — every trial capped there, and compounded 10/15/30 mg vials are containers for sub-2.4 mg dosing, not a license to exceed the envelope.
What Semaglutide Is
A modified version of the body's own GLP-1, with a fatty-acid side chain that binds albumin to extend half-life so one injection covers ~a week. GLP-1 is released after eating: it signals the brain that food has arrived, slows gastric emptying, and coordinates insulin/glucagon. Semaglutide amplifies and prolongs that signal. Forms: once-weekly injection (Ozempic/Wegovy) and daily oral tablet (Rybelsus, using an absorption enhancer).
How Semaglutide Works
| Receptor | What it does | Semaglutide |
|---|---|---|
| GLP-1R | Appetite suppression, gastric slowing | 1.0× |
| GIPR | Insulin efficiency, fat metabolism | — |
| GCGR | Liver fat oxidation, energy expenditure | — |
As a pure GLP-1 agonist at full native potency, it works through appetite, fullness, gastric slowing, and glucose coordination — quieting hunger and food-seeking, increasing fullness, reducing intrusive food thoughts, and smoothing post-meal glucose. Two consequences: GI side effects are common (same pathway slows the gut), and without GIP its non-diabetic obesity DXA readout is less fat-selective (~62:38 fat:lean vs tirzepatide's ~75:25).
What the Trials Measured
STEP-2 (the direct dose comparison): the only GLP-1 trial directly comparing 1.0 vs 2.4 mg for weight loss (BMI ≥27 + T2D, 68 weeks).
| Arm | Weight loss at 68 wk | Δ vs placebo | % of 2.4 mg effect |
|---|---|---|---|
| Placebo | −3.4% | — | — |
| 1.0 mg/wk | ~7.0% | +3.6 pp | ~73% |
| 2.4 mg/wk | ~9.6% | +6.2 pp | 100% |
STEP-1 (non-diabetic anchor): ~15% mean weight loss over 68 weeks at 2.4 mg (half reached ≥15%, a third ≥20%). STEP-5: durable maintenance to 2 years. SELECT (~4 yr, n=17,604): 20% MACE reduction and the largest/longest GLP-1 safety dataset. FLOW (1.0 mg): kidney-outcome benefit in T2D + CKD. SUSTAIN-FORTE: 1.0 vs 2.0 mg for glycemic control (~1.9% vs ~2.1% HbA1c). Body composition (STEP-1 DXA): ~62:38 fat:lean — without training/protein, ~40% of lost mass can be lean; in T2D the gap with tirzepatide narrows (Clamp study ~86:14 vs ~87:13). SURMOUNT-5 (head-to-head): semaglutide 13.7% vs tirzepatide 20.2% at 72 weeks. Semaglutide's edge is cardiovascular outcomes and oral availability, not raw magnitude.
Dosing
Once weekly, subcutaneous. Four ideas carry the decision: titration-is-not-therapeutic, stop-short, microdose, and the body-weight-exposure effect.
Titration ≠ therapy: the 0.25 mg starter is non-therapeutic GI adaptation (the first effective dose is 0.5 mg); each step holds ≥4 weeks (the ~7-day half-life reaches steady state in 4–5 weeks). Stop-short: 0.5 mg (well-characterized T2D maintenance across SUSTAIN) and 1.0 mg (~73% of 2.4 mg effect per STEP-2) are defensible maintenance doses; the effective dose is the lowest that keeps weight trending. The exception is cardiovascular protection — the 20% MACE reduction was measured at 2.4 mg. Microdose (0.1–0.25 mg): below the 0.25 mg/wk floor is uncharacterized in weekly trials; the case rests on a daily-SC Phase 2 anchor (~0.35 mg/wk equivalent) plus linear PK — the thinnest microdose evidence of the three modern GLP-1s.
Body weight shifts exposure (unique to semaglutide):
| Body weight | Exposure vs 85 kg reference |
|---|---|
| 55 kg | +40% |
| 70 kg | +15% (approx) |
| 85 kg | reference |
| 100 kg | −10% (approx) |
| 127 kg | −27% |
Lighter users may respond at lower milligrams; heavier users may need to climb further. Oral-SC equivalence (correcting a common error): 14 mg oral daily ≈ ~0.5 mg SC weekly (not the commonly cited 1.0 mg). Rybelsus must be taken ≥30 min before food with ≤4 oz plain water — the strictest fasting requirement in the class.
| Band | Dose | Evidence grade |
|---|---|---|
| Microdose | 0.1–0.2 mg | Mechanism-consistent (below trial coverage) |
| Titration floor | 0.25 mg | Directly measured — explicitly non-therapeutic (4 weeks) |
| T2D maintenance | 0.5 mg | Directly measured (SUSTAIN-1 to 11) |
| Stop-short maintenance | 1.0 mg | Directly measured (STEP-2; ~73% of 2.4 mg) |
| Full weight-loss dose | 2.4 mg | Directly measured (STEP-1/2, SELECT); full CV benefit |
| Above 2.4 mg | — | Uncharacterized — do not exceed |
Side Effects
Primarily gastrointestinal, dose-dependent, worst during the first 4–8 weeks. Common (STEP-1): nausea 44%, diarrhea 30%, vomiting 25%, constipation 24%, abdominal pain 20%. Less common: fatigue, headache, dizziness, reflux, weight-loss-associated hair thinning. Serious but rare: gallbladder problems (1–3%, from rapid weight loss), pancreatitis (<1%), thyroid contraindication (personal/family MTC or MEN2), hypoglycemia (mainly with insulin/sulfonylureas). Management: slow titration (4+ weeks/step), smaller protein-first meals, avoid high-fat/greasy foods, hydrate, and hold at the prior step if a reaction hasn't resolved rather than layering a new dose on top.
Cost: Brand vs Compounded
| Aspect | Brand (Ozempic/Wegovy) | Compounded |
|---|---|---|
| FDA approval | Yes | No |
| Manufacturing | Novo Nordisk | Compounding pharmacies (503A/503B) |
| Consistency | Standardized | Variable by pharmacy |
| Delivery | Pre-filled pens | Usually vials + syringes |
| Cost | $900–1,700/mo | $200–500/mo |
If using compounded, do so through a clinician who takes responsibility for the supply chain and can verify quality (certificate of analysis, 503B oversight).
Where Semaglutide Is Evidence-Led
Cardiovascular protection (SELECT — the only GLP-1 with placebo-beaten CV outcomes in non-diabetics); MASH histologic resolution (ESSENCE — first positive histologic MASH endpoint for an incretin); direct dose-comparison evidence (STEP-2, SUSTAIN-FORTE); safety-record duration; low immunogenicity (1–3% anti-drug antibodies vs tirzepatide's 51–64%); the only GLP-1 with an oral form; and the widest insurance coverage. It produces smaller absolute weight loss than tirzepatide (~15% vs ~21%) and retatrutide, but it's the evidence-led choice when CV risk, MASH, oral delivery, chronic-safety characterization, or immunogenicity is the priority.