GLP-1 microdosing: what people mean by it, and what is actually known

Both Wegovy and Zepbound begin below the dose they are meant to settle at, because starting low reduces nausea, and both labels say in so many words that the starting dose is not approved as a maintenance dose ^[4].

• Wegovy: 0.25 mg a week for four weeks, then steps to 2.4 mg ^[4].
• Zepbound: 2.5 mg a week for four weeks, then steps of 2.5 mg to 5, 10 or 15 mg ^[4].
• Ozempic: 0.25 mg for four weeks, then 0.5, 1 or 2 mg as blood-sugar control requires ^[4].

So a dose below the starting dose has no place on any label, and staying on the starting dose forever is off-label too. The labels do allow a prescriber to hold a patient at a lower maintenance dose when a higher one is not tolerated; Zepbound's says to consider a lower maintenance dose if the patient does not tolerate one ^[4]. That is a prescriber's decision, not a dosing scheme.

On intervals, the labels give limits: a Wegovy dose can be taken within five days of the missed day, a Zepbound dose within four, and there are minimum gaps between doses ^[4]. Those are rules for a missed dose, not permission to stretch the schedule.

Two things, and neither is a trial of microdosing.

A 2026 review of the practice. It defines microdosing as 25% to 50% of the recommended starting dose or extended intervals, describes how people do it, and concludes that high-quality clinical evidence is essentially absent. Its stated worries are compounded products, non-standard dose measurement and dosing accuracy ^[1].

A large records study. Researchers looked through 29 million patient records for people who stayed on the lowest approved dose for at least six months: 814 on semaglutide 0.25 mg and 1,016 on tirzepatide 2.5 mg. Average weight change at 12 months was 2.2% on semaglutide and 5.5% on tirzepatide ^[2]. The authors are explicit that a prescription record is not proof of intentional microdosing, and the two groups were not randomized, so the comparison between them is not a clean one.

What the second study does suggest is the scale of what a sustained lowest dose achieves: a couple of percent on semaglutide over a year, against about 15% at the full weight dose in Wegovy's trials ^[4]. Several 2025 and 2026 commentaries in diabetes and obesity journals discuss the practice in multi-dose pens and vials, and all of them note the same evidence gap ^[3].

This is the part that gets least attention and causes most harm.

"Click counting" means drawing part of a dose from a multi-dose pen by counting the clicks of its dial. The pens are built to deliver one fixed dose per injection; they are not metered for fractions, and the maker has not tested accuracy that way ^[1]. Drawing a part dose from a vial with a syringe is more controllable, but depends on the person's technique and on the vial's concentration, which differs between presentations: a Zepbound multi-dose vial at the 15 mg strength holds 25 mg of drug per mL, six times the concentration of the 2.5 mg strength ^[4]. Getting the arithmetic wrong on a concentrated vial means a large overdose, not a small one.

Compounded products add a second unknown: the strength is whatever the pharmacy made it. See compounded versus branded.

Side effects are dose-related, so a lower dose generally means less nausea; that part of the rationale is sound. But the effects people most want to avoid are mostly during escalation and ease over time, as our side effects page sets out.

The serious risks on the label are not dose-dependent in a way anyone has mapped: pancreatitis, gallbladder disease and allergic reactions are not listed as applying only above a certain dose ^[4]. A smaller dose is not a dose with no risk, and the warnings apply to anyone taking the drug.

The approved benefits belong to the doses that were tested. The weight numbers, the heart-outcome result and the sleep-apnea result all come from trials of the maintenance doses ^[4]. A quarter dose has none of that behind it.

The honest framing is that this is an off-label experiment with no published results. If cost is the reason, the cash prices and programs in how to get a GLP-1 are worth going through first, including the $50 Medicare route for people who qualify. If nausea is the reason, holding the dose or stepping down is something the label anticipates and a prescriber can do properly ^[4].

Whatever you decide, do it with the prescriber rather than around them: the dose in your record is what any other clinician will assume you are taking.