GLP-1 microdosing: what the evidence does and does not show

What people mean by microdosing

There is no standard medical definition. In news reports, telehealth companies sell "microdosed" programs that use compounded semaglutide or tirzepatide at doses below those of the approved products. STAT reported that companies make claims such as lower metabolic risk, lower inflammation and a lower risk of cognitive decline, and the Washington Post, as relayed by ConsumerAffairs, counted at least 15 telehealth companies and practices promoting it for longevity.

Online posts describe several versions of the idea. None of it was measured, and the Atlas does not repeat amounts, because two people who both say "microdosing" can mean different things.

What the labels say about starting doses

A prescriber can adjust an approved product, for example by delaying a dose increase or lowering a maintenance dose. That is a change made inside the label's framework. It is not the same as a plan to stay at a fraction of the starting dose.

This describes label structure for information and gives no amounts. The schedules are on the labels and are for a prescriber to apply. In the dosage sections the Atlas read, neither label uses the term microdosing.

What trials show

The Atlas found no randomized trial of microdosing semaglutide or tirzepatide for weight loss. A ClinicalTrials.gov search on October 23, 2026 for the word microdosing in titles, summaries and keywords, combined with semaglutide or tirzepatide, returned no registered studies.

One registered study has "low-dose" in its title, a single-arm pilot of 60 adults in Pakistan (NCT07513168). Its registration describes a standard step-up schedule over 24 weeks with lifestyle counseling, not a fraction of a standard dose, so it is not a test of microdosing. It was recruiting and had no results.

The closest controlled evidence is about lower doses after weight loss, not at the start. In SURMOUNT-MAINTAIN, people who moved to a lower labeled tirzepatide dose were 16.6% below their starting weight at week 112, compared with 21.9% for those who kept the highest dose they tolerated and 9.9% for placebo. The regain page covers it. It answers a different question, because those people had first lost weight on a full course.

The closest data: studies that compared dose bands

A 2026 analysis of electronic health records from a U.S. network (the authors work for the data company nference) compared matched people who had used semaglutide with people who used metformin, and looked at COVID-19 and influenza severity. Semaglutide users had lower 30-day mortality, hospitalization and a combined severity measure. Severity did not differ across the semaglutide dose bands the authors compared, and they wrote that the result motivates prospective evaluation of low-dose semaglutide.

That is a hypothesis, not a test of microdosing. The dose bands were label strengths people were on, not a planned microdosing regimen, and the comparison was observational, so it cannot show that a small dose works or is safe.

What the FDA and the makers say

• The FDA's page on unapproved GLP-1 drugs does not use the word microdosing in the sections the Atlas read. Its dosing concerns are dosing errors and doses, frequencies or titration schedules beyond the approved label. It urges prescribers to be vigilant when setting doses and titration, and patients to confirm with their prescriber or compounder how the intended dose is measured and given.
• The FDA's July 2024 alert describes patients unfamiliar with syringes, confusion between units, milliliters and milligrams, and clinician errors converting between them. Some patients were hospitalized.
• ConsumerAffairs reports that Eli Lilly said it has no data on the benefits or risks of microdosing Zepbound or Mounjaro, and that its vials are single-use and contain no preservatives. It also relays a quote from Daniel Drucker of the University of Toronto, whose research microdosing supporters cite, that it is uncertain whether low doses reduce inflammation. Both are secondary reports.
• STAT quotes physicians and researchers who say strong clinical evidence for very small doses is lacking. A Yale clinician-researcher told STAT that people using microdosed products are in effect test subjects for both effectiveness and safety. Part of that article is behind a paywall.

Claims made for microdosing, and what the Atlas found

Why measuring comes up

Any plan that involves taking less than a labeled dose raises the question of how the amount is measured and who checks it. In a 2023 poison center series, two of three people made ten-fold self-administration errors with compounded semaglutide from vials, and one dosed in milliliters and units instead of milligrams. A 2026 case report describes a person who needed four days of breathing support after severe swelling of the airway and later reported a ten-fold dosing error. The page on FDA safety reports covers what the reports show.

The Atlas does not give measuring or mixing instructions. A clinic or pharmacy that supplies anything that needs measuring should teach it directly.

Quick answers

What is GLP-1 microdosing?

There is no standard definition. It usually means using a small fraction of the labeled dose of semaglutide or tirzepatide, often compounded, for goals such as appetite control, lower inflammation or longevity.

Does microdosing work for weight loss?

Not shown. The Atlas found no randomized or registered trial of it, and no label that describes it.

Is microdosing safe?

It has not been tested. The FDA has reported dosing errors and doses outside the approved labels with compounded products, and the published cases include ten-fold errors.

Is microdosing the same as a starting dose?

No. A starting dose is the first step in a rising schedule. The Zepbound label says its starting strength is for treatment initiation and is not approved as a maintenance dosage.

Do Wegovy or Zepbound have a microdose?

In the dosage sections the Atlas read, neither label uses the term. They describe stepped schedules and say what to consider when a dose is not tolerated.

Have any trials tested microdosing?

None that the Atlas found. A ClinicalTrials.gov search on October 23, 2026 returned no registered study that uses the term.

Why are compounded products part of this?

The news reports the Atlas read describe microdosing programs that use compounded products. Compounded products are not reviewed by the FDA before they are sold.

Do medical societies recommend it?

The Atlas found no society guideline that recommends microdosing. Manufacturer and researcher views here come from news coverage.

Questions to ask

• Is there a published study behind the dose you are proposing, and what does the approved label say about it?
• Is this product FDA-approved or compounded, and who makes it?
• How will the amount be measured, and who will show me?
• What are we trying to achieve, and how will we know in a few months whether it is working?
• What is the plan if side effects appear or weight loss stalls?

What is still unknown

• The Atlas found no randomized trial of microdosing semaglutide or tirzepatide, and no registered trial that uses the term as of October 23, 2026.
• Microdosing has no standard definition, so reports about it describe different things.
• Whether small doses work for weight, inflammation, longevity or cognition is untested.
• Whether staying on small doses causes fewer side effects than the stepped schedules is untested.
• How many people microdose, and how they get and measure the medicine, was not measured in the sources the Atlas found.
• Manufacturer and researcher views reported here come from news coverage, which the Atlas treats as secondary.