Every dose on an approved label is the output of years of work. Understanding how those numbers are produced explains why dosing cannot be reasoned out from first principles, and why an unapproved compound has no dose at all.
Finding the range
Phase 1 trials typically use dose escalation: a small group receives a low dose, safety is assessed, and the next group receives more. This maps the range between no measurable effect and unacceptable side effects.
Retatrutide’s first human study in 2019 gave 45 healthy volunteers single doses between 0.1 mg and 6 mg. It also established a half-life of about six days, which is why later trials dosed weekly.
Choosing between doses that work
Phase 2 compares several doses for effect. The point is rarely to find the strongest dose; it is to find the best trade-off. In retatrutide’s phase 3 obesity trial the results were 19.0% weight loss on 4 mg, 25.9% on 9 mg and 28.3% on 12 mg, while discontinuation for side effects rose from 4.1% to 11.3% across the same range.
Why escalation schedules exist
Nearly every GLP-1 class product starts below its maintenance dose and increases in steps. Wegovy runs 0.25 mg for four weeks, then 0.5, then 1, then 1.7, reaching 2.4 mg at week 17. Zepbound starts at 2.5 mg, explicitly described in its labelling as an initiation dose that is not approved for maintenance.
The reason is tolerability. Gastrointestinal side effects cluster immediately after each increase, and stepping slowly keeps more people on treatment.
Dose is tied to population and formulation
- Population: tesamorelin’s 1.4 mg daily applies to HIV-associated lipodystrophy. It is not a general dose.
- Kidney function: elamipretide drops from 40 mg to 20 mg daily in adults with significantly reduced kidney function.
- Route: semaglutide’s injectable and oral forms use entirely different numbers.
Why study protocols are not recommendations
Trial doses were given with pharmaceutical-grade material, under medical supervision, to screened participants, with monitoring and a route to stop. Removing all four of those conditions does not leave the dose intact.
Where no human study has established a dose, no dose exists to report. Pages on this site say so rather than estimating one, because an estimate would be invention.