Microdose GLP-1: what the trial data actually shows

The microdose GLP-1 conversation has been gaining ground since late 2024 in the longevity-adjacent corners of Twitter and a few Substacks. The pitch: take a fraction of the standard dose, get most of the metabolic benefit (insulin sensitivity, inflammation, appetite quiet) with fewer side effects and a much lower spend. The idea is intuitive enough that it's now showing up in actual telehealth menus.

The data is messier than the pitch.

What "microdose" actually means here

There's no shared definition. In the discourse it ranges from 0.05mg semaglutide weekly (about 1/50th of a maintenance dose) up to about 0.25mg, which is a quarter of the lowest commercial Wegovy dose. For tirzepatide it's roughly 0.5mg to 1.25mg weekly, where 5mg is the typical low maintenance dose.

This range is wide enough that two people both saying "microdose" might be on protocols that differ by 5x. When you read someone's experience report, ask which they meant.

The trial data in one paragraph

There is no completed Phase 3 trial of microdose GLP-1 for weight loss or metabolic outcomes. The original semaglutide STEP-1 and tirzepatide SURMOUNT-1 trials tested doses ranging from 0.5mg semaglutide and 5mg tirzepatide upward. The lower-dose arms in those trials were dose-escalation phases for the comparator dose, not standalone treatment arms. The closest analog is some of the type 2 diabetes work where 0.25mg semaglutide weekly was a maintenance dose for some patients, with measurable but smaller HbA1c improvements than 1mg.

What this means in practice: the question "does 0.1mg semaglutide do something useful at a population level" hasn't been answered in a controlled trial. Anecdotes exist. Mechanistic reasoning exists. RCT evidence does not.

What the mechanism does support

GLP-1 receptor occupancy isn't binary. Lower doses produce lower receptor occupancy, and lower receptor occupancy produces lower (but non-zero) downstream effects. The pharmacology of these drugs is well-characterized and there's no biological reason a 1/10th dose would do nothing. The question is whether the something it does is clinically meaningful for the specific outcome you care about.

For appetite suppression: probably some effect, much smaller than full dose, with high inter-individual variability. Some people report meaningful appetite quiet at 0.1mg semaglutide. Others report nothing.

For glucose control: subtle effects at low doses are documented in T2D populations, but the clinical impact at sub-therapeutic doses in non-diabetic populations is not well-characterized.

For inflammation markers and cardiovascular risk: this is where the longevity case gets made, but the trial data on cardiovascular benefit is at full doses (SELECT trial used 2.4mg semaglutide). Whether 1/10th the dose gives 1/10th the benefit, or 1/2 the benefit, or no benefit, is genuinely unknown.

For weight loss alone: small. Most people who lose meaningful weight on microdose are also doing other things (exercise, sleep, calorie awareness) that the GLP-1 isn't doing the heavy lifting on.

Who's actually prescribing this

A growing slice of US compounding-focused telehealth is now offering microdose protocols. Some Asia-Pacific clinics in Singapore and Bangkok have followed. The framing is usually metabolic health and longevity rather than weight loss specifically, which is the right framing given the data.

A small number of endocrinologists will write microdose protocols off-label for patients who tolerated maintenance doses poorly but want to keep some metabolic effect. This is medically defensible and the dosing is conservative.

What you should be wary of is anyone selling microdose as "all the benefits with none of the side effects." That isn't supported. Microdose has fewer side effects because it has less drug effect; the trade-off is real, not free.

The honest read

If you're metabolically healthy, want to try a low dose to see how your body responds before committing to a full course, and you can afford the risk of spending money on something that might do less than the marketing implies, microdose is a reasonable experiment. Track something objective (fasting glucose, weight, energy patterns) and reassess at twelve weeks.

If you have meaningful obesity or T2D and a doctor would prescribe full-dose GLP-1, microdose is almost certainly suboptimal. The trial data supports the doses that were tested. Don't undershoot for ideological reasons.

If you're chasing longevity benefits specifically, the SELECT cardiovascular outcomes data is at 2.4mg semaglutide. The longevity case for 0.1mg is mostly extrapolation, not evidence.

The thing I keep coming back to is that the absence of trial data doesn't make microdose pseudoscience, but it also doesn't make it as well-supported as the social-media discourse implies. Treat it as an experimental protocol, not as an established treatment, and the decision becomes easier to think clearly about.