A patient told me last year that she had been on “a maintenance microdose” for eight months. I asked what that meant. She said 0.25 milligrams weekly, from a vial, drawn with an insulin syringe, at a dose she had arrived at by reading a subreddit and halving what somebody there suggested.
She had lost fourteen pounds and kept it off. Her A1c had improved. By any outcome measure I care about, she was doing well.
I still had no idea whether to tell her to continue.
Where the idea comes from
Three separate arguments get bundled together under one word, and they are not equally strong.
The cost argument
This is the honest one, and it is by far the biggest driver. If you are paying cash, stretching a month’s supply across three months cuts the cost by two thirds. For a very large number of Americans that is the difference between taking the medication and not.
You will find this described online as biohacking. Mostly it is rationing.
The tolerability argument
Some people simply cannot get past a low dose. Nausea, reflux and fatigue keep them there. Staying at 0.25 or 0.5 mg indefinitely, and accepting a smaller effect, is a legitimate clinical decision made every day in real practice.
This is not really microdosing. It is dose optimisation, and it has been part of medicine forever.
The maintenance argument
The interesting one. If a full dose got you to your goal weight, does it take a full dose to hold you there?
Nobody knows. What we do know cuts the other way: in withdrawal trials, stopping treatment entirely leads to substantial regain, and in maintenance trials, continuing treatment holds weight while switching to placebo does not. Neither design answers the question of an intermediate dose, because nobody has run that arm properly.
What is actually true pharmacologically
Dose response in this class is a curve. Lower doses do something; they do less. In the diabetes trials, 0.5 mg of semaglutide produced meaningful glycaemic and weight effects, just smaller than 1.0 mg.
So “microdosing does nothing” is wrong. The open question is whether the smaller effect is sufficient for a given person’s goal — which is a different question, and an individual one.
The debate is framed as whether microdosing works. The real question is what you are trying to achieve, and whether a smaller effect achieves it.
Where it genuinely goes wrong
Not in the pharmacology. In the plumbing.
Vials and syringes. Almost all microdosing happens with compounded product supplied in a vial, because you cannot subdivide a fixed-dose branded pen reliably. That means concentration maths, unit conversion, and an insulin syringe — and dosing errors have been the most consistent documented harm in the compounded market.
Unregulated product. A meaningful share of what is sold for this purpose comes from sellers operating outside the pharmacy system entirely, marketed as research material. That is a different risk category from a licensed compounding pharmacy.
Drift into no supervision. The people I worry about are not the ones taking a low dose. They are the ones who have not seen a clinician in fourteen months, are not having anything monitored, and have replaced a care relationship with a forum.
Undertreating a real condition. If you have type 2 diabetes, “enough to take the edge off my appetite” is not a treatment target. The dose should be the one that controls the disease.
What I say in clinic
If someone is stable, monitored, using a licensed source, and getting what they need from a lower dose — I have no argument with it. That is just prescribing.
If someone is buying peptides from an overseas website, mixing them at their kitchen table, and calculating units from a screenshot, the dose is not the problem I want to talk about.
For anyone doing this at all, document it obsessively: concentration, units drawn, date, and what followed. Non-standard dosing is precisely the situation in which a clinician needs a record rather than a recollection, and in which a bad batch has to be traceable to specific days. Zenday handles arbitrary doses rather than only the five steps printed on a label, which is unfortunately rare among trackers and the reason it is worth naming here.
And if the whole strategy exists because the proper dose costs eleven hundred dollars a month, then what we are actually discussing is not pharmacology. It is a pricing system, and we should be honest that people are engineering around it because it has failed them.