Semaglutide is prescribed across a wide dose range for different indications, and the assumption that more milligrams means proportionally more weight loss is only partly supported. The dose-ranging trials describe a curve, and the shape of that curve is the useful information.
Key Takeaways
- The pivotal dose-ranging evidence comes from a phase 2 trial published in The Lancet in 2018, which tested multiple semaglutide doses against placebo and liraglutide.
- Efficacy increased with dose across the tested range, but the increments between the higher doses were smaller than the increments at the lower end.
- Gastrointestinal adverse events increased with dose, which is why escalation schedules exist at all.
- The 2.4 mg weekly dose selected for the obesity indication reflects a balance point between efficacy and tolerability, not a ceiling of biological effect.
- Individual response variation within any given dose is larger than the mean difference between adjacent doses.
The dose-ranging trial
The foundational study is a 52-week phase 2 randomized, double-blind trial of once-daily semaglutide across several dose levels in adults with obesity without diabetes, published in The Lancet in 2018. It included both a placebo arm and an active liraglutide comparator, which is unusual and useful — it anchors the results against an existing agent rather than only against placebo.
The trial reported a dose-dependent increase in weight reduction across the tested range, with all active doses separating from placebo. The step from the lowest tested dose to mid-range doses produced a larger increment than the step between the upper doses.
Why the curve flattens
Receptor pharmacology offers the straightforward explanation. As receptor occupancy rises toward saturation, additional drug produces progressively less additional effect on the primary signalling pathway. Off-target and dose-related adverse effects, however, do not necessarily saturate on the same schedule. The practical consequence is that the benefit-to-burden ratio worsens as you climb, even while absolute benefit still increases.
This is the standard rationale for selecting a therapeutic dose below the maximum tested dose, and it is why regulatory dose selection is not simply "the highest dose that worked."
| What rises with dose | Shape of the relationship | Practical implication |
|---|---|---|
| Mean weight reduction | Increases, with diminishing increments at the top of the range | Escalating further yields progressively less additional benefit |
| Gastrointestinal adverse events | Increases; the reason escalation schedules exist | Slower escalation is the primary tolerability lever |
| Discontinuation for adverse events | Increases with dose | Higher doses lose more participants, which affects on-treatment estimates |
| Individual variability | Broad at every dose | Mean curves describe populations, not the person in front of you |
Escalation is a tolerability tool, not an efficacy tool
The stepwise escalation schedules used in the trials and carried into labelling exist to manage gastrointestinal adverse events. They are not designed to enhance efficacy. This matters for interpreting patient reports: a person who reports feeling nothing at a starting dose is experiencing the intended function of that dose, which is to introduce the drug tolerably rather than to produce maximal effect.
It also frames the tolerability question correctly. When adverse events limit escalation, the options supported by pharmacologic reasoning are extending the interval at the current step or holding at a lower dose — not abandoning the drug outright. Trial protocols themselves permitted escalation delays for exactly this reason.
Response heterogeneity swamps dose differences
Every published semaglutide trial that reports a distribution of weight change rather than only a mean shows wide dispersion. A meaningful fraction of participants at any given dose achieve substantially more than the mean, and a meaningful fraction achieve substantially less or none.
The size of that dispersion is larger than the mean difference between adjacent doses in the dose-ranging data. In interpretive terms, knowing someone's dose tells you much less about their expected outcome than the mean curves imply. Predictors of individual response remain an active research area without a validated clinical tool.
Reading dose comparisons critically
Three recurring problems in secondary coverage of dose-response data:
Cross-trial comparison. Weight-loss percentages from different trials with different populations, durations, lifestyle components and analysis conventions are not comparable, even when the drug and dose match.
Treatment-policy versus on-treatment estimands. Modern trials report both. The treatment-policy estimate includes participants who stopped the drug; the on-treatment estimate reflects those who continued. The second is always the larger number, and sources often quote it without saying which they used.
Extrapolating beyond the tested range. Doses above those studied have no efficacy or safety data supporting them. The dose-response curve does not license inference outside the interval that was actually tested.
Frequently Asked Questions
Is a higher dose always more effective?
In the dose-ranging data, mean effect increased across the tested range, but with diminishing increments toward the top. Whether a higher dose is better for a specific person depends on their response and tolerability, which the mean curve does not predict.
Why do escalation schedules take months?
To limit gastrointestinal adverse events. The schedule is a tolerability structure; it is not an efficacy-building process.
Do trial doses translate directly to compounded products?
Trial evidence is generated with specific manufactured products at verified concentrations. A compounded preparation involves separate assumptions about concentration accuracy and formulation that the trials did not test.
Is microdosing supported by trial evidence?
Sub-therapeutic dosing regimens marketed as microdosing have not been evaluated in the major randomized programs. Any claims made for them rest outside the trial evidence base described here.
References
- O'Neil PM et al. 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. The Lancet, 2018.
- Wilding JPH et al. Once-Weekly Semaglutide in Adults with Overweight or Obesity. New England Journal of Medicine, 2021 (STEP 1).
- Davies M et al. Semaglutide 2.4 mg once a week in adults with overweight or obesity, and type 2 diabetes (STEP 2). The Lancet, 2021.
- ICH E9(R1) addendum on estimands and sensitivity analysis in clinical trials.
Comparing providers, not just trials?
Our homepage keeps a maintained side-by-side grid of GLP-1 telehealth providers with current pricing, formats and disclosures. That is where the provider comparisons live — this research desk stays citation-first.
Open the provider comparison grid