What Sub-Therapeutic GLP-1 Dosing Is, and What the Randomised Trials Actually Measured

Every approved GLP-1 amount comes from a titration schedule that a randomised trial tested. Microdosing deliberately sits below it. That is the whole subject: a practice defined by its position relative to an evidence base it falls outside.

Semaglutide and tirzepatide are the best-evidenced compounds in the weight loss and metabolic peptides category, with tens of thousands of randomised participants between them, which makes the gap here unusually sharp — the drugs are extremely well studied and the practice is not studied at all. The semaglutide calculator, the tirzepatide calculator and the reconstitution calculator handle the arithmetic below. Both are prescription medicines, every amount is a prescriber's decision, and this page is a description of a practice rather than a protocol.

What is GLP-1 microdosing?

Microdosing is defined by where it sits, not by what it is. The practice means giving semaglutide or tirzepatide at an amount below the lowest rung of the approved titration schedule — below semaglutide's 0.25 mg weekly starting dose, or below the 2.5 mg increment tirzepatide is titrated in. No published definition fixes a figure, no trial has tested one, and the amounts in circulation are chosen by individuals rather than derived from anything. That is the first fact about it and it decides most of the rest.

Three features of the practice are worth separating, because they are usually discussed as one thing.

The first is arithmetic: a microdose is a fraction of a labelled amount, drawn from a vial reconstituted at a concentration the person chose. The second is a claim: that a smaller amount delivers a useful share of the effect. The third is a rationale: that a smaller amount is easier to tolerate. The arithmetic is checkable, the rationale has a real basis discussed below, and the claim has no trial behind it at any amount for either molecule.

There is also a naming point worth clearing early, because it recurs in vendor catalogues. "GLP-1SG" is a catalog code for semaglutide, not a separate compound or a lower-strength variant.

What amounts did the GLP-1 trials actually test?

Every published efficacy result attaches to a maintenance amount. Semaglutide's approved schedule starts at 0.25 mg weekly and escalates over months to a standard weight-management dose of 2.4 mg weekly, with 7.2 mg approved in March 2026; tirzepatide is titrated in 2.5 mg steps, and SURMOUNT-1 randomised 2,539 participants for 72 weeks at 5, 10 and 15 mg. The escalation rungs exist to be passed through on the way to those amounts. None of them is an endpoint a trial reported against.

Two things that table deliberately does not contain are worth naming. It carries no week-count between rungs, because the label states escalation runs over months and the specific interval is a prescribing instruction; inventing a number for it would be the exact failure this site exists to correct. And it keeps label amounts separate from trial arms, because 2.5 mg is the increment tirzepatide's label titrates in while 5, 10 and 15 mg are the arms a randomised trial ran. Those are different kinds of fact.

One further misreading closes here. A "68-week protocol" is a trial duration, not a treatment plan — STEP 1 stopped at 68 weeks because that was its design. The ladders themselves are set out rung by rung in the semaglutide dosing chart and the tirzepatide dosing guide.

Is there efficacy data below the approved amounts?

No. Not one published randomised trial has measured weight loss, glycaemic control or cardiovascular outcomes at any amount below the label ladder, for either molecule. The lowest amounts with a published efficacy result are semaglutide's 0.5 and 1.0 mg in SUSTAIN-6, which are label strengths in type 2 diabetes rather than sub-therapeutic amounts, and tirzepatide's 5 mg arm in SURMOUNT-1. Below those, the record is empty — not weak, not mixed, empty.

That absence has a structural reason worth understanding rather than lamenting. A titration schedule is not a dose-ranging study. Its purpose is to move a patient through low amounts quickly enough to reach a target while limiting adverse events on the way, so the low rungs are never held long enough or measured carefully enough to produce an efficacy estimate. A dose-ranging trial holds each amount and measures the endpoint at each one. Nobody has run that below the ladder.

Read the last row against the six above it. This is a compound class where the evidence is exceptionally good, which makes the absence at the bottom of the ladder informative rather than merely inconvenient: the trials that would have caught a sub-therapeutic benefit were run, funded and published, and they were not designed to look below the schedule because the schedule was the point. The full records sit in semaglutide: the nausea rate is 44.2%, not 73% and tirzepatide: the dual GIP/GLP-1 agonist that beat semaglutide by 6.5 points.

Where does sub-therapeutic dosing have a rational basis?

In tolerability during escalation, and the labels already implement it. Semaglutide 2.4 mg produced nausea in 44.2% of STEP 1 participants against 17.4% on placebo, concentrated during dose escalation, with 4.5% discontinuing for gastrointestinal reasons; tirzepatide's discontinuation for adverse events ran 4.3%, 7.1% and 6.2% across its three doses against 2.6% on placebo. Both labels start low and escalate over months for exactly that reason. The rationale is real, and it is a rationale for slower titration rather than for a permanent amount below the ladder.

That distinction is the substantive one on this page. Going slowly through the escalation window is what the approved schedule does; stopping inside the window and staying there is a different proposition, and it is the one with no data. The first is a prescribing decision with a labelled basis; the second is an efficacy claim nobody has tested.

The shape of that profile matters more than the headline percentage. A 44.2% incidence coexisting with a 4.5% discontinuation rate means most people who experience nausea continue, and the two numbers describe different things. The 73% figure that circulates online appears nowhere in the trial. Cross-compound context sits in the peptide side effects hub, and how fast to move between rungs is a question for a prescriber rather than for a page.

Why doesn't tolerance to nausea mean the appetite effect is preserved?

Because the two run on different systems, and this is the mechanistic finding that most directly undercuts the microdosing argument. Friedrichsen and colleagues measured ad libitum energy intake 35% lower than placebo — 1,736 against 2,676 kJ — with reduced hunger and increased satiety, and found no delayed gastric emptying at week 20. Slowed gastric emptying is a real acute effect that attenuates with continued dosing; the appetite effect runs on hypothalamic and brainstem circuits and does not.

Friedrichsen et al., writing in Diabetes, Obesity and Metabolism in 2021, is the source, and the consequence for a sub-therapeutic argument is direct. If the durable mechanism were mechanical — a slower stomach — then tolerance to the nausea would imply tolerance to the appetite suppression, and a low amount that avoided the first would be trading away the second in proportion. It does not work that way, because the two are separate systems.

What that does not establish is the microdosing conclusion. Showing that the appetite mechanism is central and durable at 2.4 mg says nothing about how much of it survives at a fraction of 0.25 mg, because nobody has measured the appetite effect at those amounts. A mechanism that is dose-dependent at the amounts tested cannot be assumed to be dose-independent below them. The honest reading is that the mechanism is understood and the dose-response curve below the ladder is not.

What does a sub-therapeutic amount measure on a syringe?

Small enough that the barrel becomes the limiting instrument. Units on a U-100 syringe are volume markings, so one unit is 0.01 mL regardless of contents, and units equal amount in milligrams divided by concentration in mg/mL multiplied by 100. At 5 mg/mL, semaglutide's 0.25 mg starting dose is 5 units; at 10 mg/mL it is 2.5 units. Half of that starting dose is 2.5 units and 1.25 units respectively, which is where graduations stop carrying information.

Worked example — the starting rung and half of it

That is an example of the arithmetic, not a recommendation to draw any of those amounts. What it demonstrates is a physical constraint rather than a pharmacological one: a 1 mL U-100 barrel is graduated in whole units, so an intended amount that lands between marks is being estimated rather than measured, and the estimate error is a large proportion of a small amount. Choosing a lower concentration at reconstitution moves the figure up the barrel, and the diluent volume is the only variable here you fully control — it is also irreversible once added.

Run both columns through the semaglutide calculator or the tirzepatide calculator alongside the reconstitution calculator before anything is mixed. The general form of the conversion, including the microgram steps and the error classes it produces, is in how to calculate peptide doses.

What happens to a small amount when the vial holds more than the label says?

The error term swamps it. Across 930 independent tirzepatide tests on this platform, vial contents run from 2.6% to 26.7% over label, with vials labelled 10 to 65 mg testing between 10.71 and 68.1 mg. On a ladder whose increment is 2.5 mg, a 26.7% overage is larger than the increment. On an intended amount below the starting rung, it is a larger proportional error than the amount being aimed at, and nothing about the liquid reveals it.

The mechanism is straightforward and it is the reason purity figures do not help here. Purity is a proportion and net peptide content is a quantity, so a vial testing 99.75% pure can still contain a quarter more drug than its label claims — and every downstream calculation inherits the label mass as an assumption. The concentration written on the vial is derived from that assumption, and so is every unit reading taken afterwards. The mechanism is set out in why 10 mg isn't 10 mg.

FDA has recorded the consequence in the direction that matters. The agency received "multiple reports of adverse events... that may be related to dosing errors associated with compounded injectable semaglutide products," resulting from patients measuring and self-administering incorrect amounts. As of 30 November 2024, FDA had received more than 392 adverse event reports for compounded semaglutide and more than 215 for compounded tirzepatide. Those are spontaneous reports rather than trial data, and the figures have not been updated since.

Has the supply route that made this practice cheap closed?

Yes, and the closure is the largest change around this practice. FDA declared the semaglutide injection shortage resolved on 21 February 2025, and enforcement discretion for compounders ended 22 April 2025 for 503A pharmacies and 22 May 2025 for 503B outsourcing facilities. FDA's shortage database now lists Tirzepatide Injection as "Resolved." Sections 503A and 503B permit compounding a drug that is essentially a copy of an approved product only while that product is in shortage, so the shortage was the permission.

Belcourt, Sapowadia & White, writing in Annals of Pharmacotherapy in September 2026, state the consequence directly: "With the shortage of innovator semaglutide or tirzepatide products resolved, compounding pharmacies can only legally sell unique products." What survives for semaglutide is narrow and patient-specific — 503A compounding under documented clinical-difference exceptions, such as a verified excipient allergy or a strength not commercially available. FDA has gone further for that molecule, proposing permanent exclusion from the 503B bulks list, which would remove the route regardless of any future shortage.

Two consequences follow for anyone reading a microdosing argument written before 2025. Cost was the whole appeal of the compounded route, and its closure is a price event as much as a legal one — what it left behind is the approved product at pharmacy prices on a prescription, or research-market vials at per-milligram prices with no prescriber, no cold chain and no regulatory recourse. There is no third option in which the compounded price survives the compounded pathway. And monitoring is not prohibition, but from 1 January 2026 markers of both molecules are monitored in and out of competition. The structural comparison is in compounding pharmacy versus research peptide, and the access question in can you get semaglutide without a prescription.

What goes wrong most often with sub-therapeutic GLP-1 dosing?

The failures cluster at the point where a small number meets an uncertain input. A units figure quoted without a concentration is unusable, because the same mark delivers different amounts from different vials; and an intended amount below the starting rung is small enough that a label overage, a between-graduation draw and a transcription slip each change it by a large proportion. The errors below are the ones with named consequences, and the first two account for most of the reported harm.

• Reading units as milligrams. A unit is a volume marking. Treating "2.5 units" as "2.5 mg" is a hundredfold class of error, and FDA has named dosing errors as a mechanism behind adverse event reports on compounded products.

• Quoting or following a units figure with no concentration attached. Without the mg/mL term the figure means nothing at all, and a protocol shared as units alone cannot be reproduced safely by anyone else.

• Treating the label mass as measured. Fill variance of +2.6% to +26.7% on 930 tests means the concentration written on the vial is an assumption, and small intended amounts inherit the whole error.

• Drawing between graduations. An amount landing at 1.25 units on a barrel marked in whole units is estimated, not measured, and the error is proportionally largest exactly where microdosing operates.

• Reading an escalation rung as a maintenance amount. The low rungs of both ladders exist to be passed through; no trial reported an outcome at any of them as an endpoint.

• Carrying a 2.4 mg or 15 mg result down to a fraction of the starting dose. Every published efficacy figure belongs to the amount that produced it, and none of them was produced below the ladder.

Injection technique errors that apply to every compound — reusing a transfer needle, injecting into a reactive site, recapping a sharp — are covered in how to inject peptides. Both molecules are prescription medicines: whether any amount is appropriate, and how quickly to move between rungs, is a clinician's decision and nothing here substitutes for it.

Which GLP-1 microdosing claims survive the evidence?

One survives in a limited form and the rest do not. The claim that adverse events concentrate during escalation is true and measured — nausea ran 44.2% in STEP 1, mostly during dose escalation, with 4.5% discontinuing — which is why both labels escalate slowly. Everything past that fails: there is no efficacy data below the ladder, no pharmacodynamic measurement at those amounts, and no basis for transferring a 2.4 mg or 15 mg result to a fraction of a starting dose.

The last two rows are the practical ones. Microdosing puts the smallest intended amounts on the least certain inputs, which is the opposite of the pairing anyone would choose deliberately.

What do platform price and lab data show?

Identity is not the failure mode on these compounds; quantity is. The Purity Index records tirzepatide at 99.75% average across 930 independent tests (verified August 2026) across 227 shops selling — one of the deepest datasets on this platform — while quantity variance runs +2.6% to +26.7%, with vials labelled 10 to 65 mg testing between 10.71 and 68.1 mg. Peptigrity tracks 530 shops and 11,852 independent lab tests across 118 peptides, with 1,283 community reviews (verified August 2026).

Price is the other half of why this practice exists, and the spread rewards reading by vial size rather than by headline figure.

Semaglutide's figures are verified 10 August 2026 and tirzepatide's 9 August 2026. Those figures exclude shipping, taxes and customs, coupon codes, bulk tiers, multi-vial kits and account-gated pricing. Vial size drives most of the visible spread, because packaging and testing costs are fixed regardless of fill — which is also why the cheapest per-milligram vial is often the one whose concentration suits a small intended amount least. Trust scores weight community reviews and independently verified HPLC purity equally at 50% each.

Read the price spread as a risk signal rather than a bargain. Both molecules have an approved version that exists on prescription, manufactured under pharmaceutical controls with a cold chain attached, and a per-milligram figure far below the rest of the market prices the absence of those things rather than discounting their presence. Individual certificates sit in the lab test database, and vendor-level detail on the semaglutide compound page.

How do you check a sub-therapeutic dosing claim?

Five checks, and the first is about the claim rather than the vial. Ask which trial measured an endpoint at the amount being recommended — not at any amount, at that one — and for both molecules the answer below the label ladder is none. The remaining checks are about whether the amount you intend is the amount delivered: net peptide content, the concentration written on the vial, a barrel that can resolve the volume, and a reverse calculation that catches transcription errors the forward one cannot see.

The first three happen before the vial is yours and the last three happen in your kitchen; neither set substitutes for the other. Vendor-level walkthroughs are in where to buy semaglutide, and anything about whether either drug is appropriate for a given person, at any amount, is a clinician's question.

The trial that would settle this

The trial that would settle microdosing is an ordinary dose-ranging study run below the approved ladder, and it has never been done for either molecule. It would hold each sub-therapeutic amount rather than escalating through it, randomise against placebo and against the approved maintenance dose, and measure weight and glycaemic endpoints at each rung with adverse events reported by arm. Without it, "a smaller amount works nearly as well" and "a smaller amount does nothing" are both untested propositions.

Frequently Asked Questions

Does microdosing semaglutide or tirzepatide work?

No published randomised trial has measured any efficacy endpoint below the approved titration ladder for either molecule, so there is no evidence-based answer. Every published result — STEP 1's 14.9% at 2.4 mg, SURMOUNT-1's 15.0% to 20.9% at 5 to 15 mg — belongs to the amount that produced it. Absence of data is not evidence that a smaller amount fails; it means nobody has looked.

What is the smallest amount with any published human data?

For semaglutide, 0.5 mg weekly appears as a SUSTAIN-6 arm in type 2 diabetes, and 0.25 mg weekly is the label's starting dose in an escalation phase rather than a maintenance amount. For tirzepatide, 5 mg is the lowest arm in SURMOUNT-1, and 2.5 mg is the increment the label titrates in. Nothing below those has a published efficacy result.

Is microdosing a way to avoid nausea?

Adverse events do concentrate during escalation — nausea ran 44.2% in STEP 1 against 17.4% on placebo, with 4.5% discontinuing for gastrointestinal reasons — and both labels escalate slowly for that reason. Moving more slowly through the schedule is a prescribing decision with a labelled basis. Remaining permanently below the schedule is a different proposition and has no data behind it.

If nausea fades, does the appetite effect fade too?

No, and this is measured. Friedrichsen and colleagues found ad libitum energy intake 35% lower than placebo, 1,736 against 2,676 kJ, with no delayed gastric emptying at week 20. Slowed gastric emptying is an acute effect that attenuates; the appetite effect runs on hypothalamic and brainstem circuits and does not. The two are separate systems.

Can you still buy compounded semaglutide or tirzepatide for this?

Largely no, for both. FDA declared the semaglutide injection shortage resolved on 21 February 2025 and enforcement discretion ended 22 April 2025 for 503A pharmacies and 22 May 2025 for 503B outsourcing facilities, while FDA's shortage database now lists Tirzepatide Injection as resolved. What survives is narrow patient-specific 503A compounding for semaglutide and, in the peer-reviewed summary's words, "unique products" only.

Why is a very small dose hard to measure?

Because a U-100 barrel is graduated in whole units of 0.01 mL each. At 10 mg/mL, semaglutide's 0.25 mg starting dose is 2.5 units and half of it is 1.25 units, which sits between marks and is therefore estimated rather than measured. Vial fill variance of +2.6% to +26.7% across 930 platform tests adds a second uncertainty on top of the first.

Where this leaves microdosing

Microdosing is a practice with a coherent motivation and no trial behind it. The motivation is real and measured: nausea ran 44.2% in STEP 1 and clustered in the escalation window, which is why both approved labels start low and climb over months. What does not follow is the efficacy claim. Every published GLP-1 result — 14.9% at 2.4 mg, 20.9% at 15 mg — attaches to a maintenance amount a titration schedule was designed to reach, and no randomised trial has measured any endpoint below the ladder.

Two practical facts sharpen that. The smallest intended amounts sit on the least certain inputs: at 10 mg/mL a fraction of the starting dose lands between graduations on the barrel, while fill variance across 930 independent tests runs +2.6% to +26.7%, so the error can exceed the amount. And the supply route that made the practice cheap has closed — both shortages resolved, enforcement discretion ended in spring 2025, and what remains is a prescription or a research vial with no prescriber attached. The compounds themselves are the best-evidenced on this platform. The practice is simply not among the things that evidence covers, and whether any amount is appropriate for any individual belongs to a clinician.

Browse the weight loss and metabolic peptides category, or our complete peptide guide with 118 compounds (verified August 2026). For per-injection volume, use the semaglutide calculator or the tirzepatide calculator alongside the reconstitution calculator. Compare shops through independent lab tests and community-verified shop reviews.

This article is for educational and informational purposes only and does not constitute medical advice. Peptides discussed may be investigational compounds not approved by the FDA (or equivalent regulators in your jurisdiction) for human use. Always consult a qualified healthcare provider before using any peptide or research compound. Peptigrity is an independent review platform and does not sell, endorse, or recommend specific products or vendors.