About one in seven people taking a GLP-1 drug in the United States is using less than the prescribed dose, according to a December 2025 analysis of nearly 76,000 users in Evidation's cardiometabolic cohort [1]. The practice is being sold as "microdosing," and the biggest direct-to-consumer telehealth platforms - Noom, Found, Hims & Hers - all launched microdose GLP-1 programs inside a three-month window in late 2025 [2]. A weight-loss physician writing in STAT News last week put it bluntly: "Microdosing GLP-1s is not a thing. There are no legitimate long-term data to support it. In fact, there isn't even a single definition of microdosing for weight loss or any other condition" [3].
What microdosing is, in 2026, is a marketing term applied to a dose that is below the FDA-approved therapeutic range, usually compounded semaglutide, usually bought through an online questionnaire. It is not a clinical protocol, and it is not a longevity protocol. It is a workaround for cost, side effects, and access, and the science has not caught up to any of those goals.
What the approved doses actually look like
For weekly semaglutide (Wegovy) the FDA has approved six doses: 0.25, 0.5, 1, 1.7, 2.4, and 7.2 milligrams. For weekly tirzepatide (Zepbound) the approved range is 2.5, 5, 7.5, 10, 12.5, and 15 milligrams. The oral form of semaglutide comes in 1.5, 4, 9, and 25 milligram daily doses [3]. The landmark STEP and SURMOUNT trials - the basis for every safety and efficacy claim these drugs carry - used only the upper end of those ranges: 1.7 or 2.4 mg of semaglutide, and 5, 10, or 15 mg of tirzepatide [3][4].
In the head-to-head SURMOUNT-5 trial published in the New England Journal of Medicine, tirzepatide produced a greater mean weight loss at 72 weeks than semaglutide at its maximum tolerated dose [5]. That is the benchmark microdose programs are now claiming to approximate, and the dose gap between "a microdose" and "the dose that produced those results" is roughly five- to ten-fold.
It is worth drawing a line that the STAT physician essay draws carefully: a prescriber who counts the clicks on a multidose Ozempic or Zepbound pen to give a patient an intermediate dose, because that patient is having trouble with side effects or losing weight faster than expected, is not microdosing. That is dose individualisation, conducted under medical supervision, and it has clinical reasons. What telehealth platforms are selling, by contrast, is a sub-therapeutic starting protocol that rarely titrates up, with a monthly price tag of $79 on the low end [6].
Why people are doing it anyway
The Evidation analysis found that microdosing is "largely patient-driven, and that there are multiple relevant influences driving microdosing decisions" [1]. Science News, summarising the same dataset, identified three overlapping motives: cost, side effects, and longevity [7]. A KFF poll reported that more than 1 in 8 U.S. adults are taking a GLP-1, and roughly half of those users describe the drugs as difficult to afford [8]. An industry analysis by HM Academy estimated annual out-of-pocket costs often exceed $3,000 and can top $4,000 at maintenance pricing, with as many as 48 million Americans expecting to start a GLP-1 in 2026 [9].
If you cannot afford the full dose, a half-dose costs half as much. The arithmetic is simple. The problem is the biology.
Katy Williams, a bariatric medicine specialist at the University of Missouri Health Care, summarised the evidence base to Science News with one sentence: "There is no rigorous scientific data to support microdosing" [7]. The compounds have been studied at their approved doses in tens of thousands of patients across the STEP, SURMOUNT, and SELECT programmes. They have not been studied at a quarter or an eighth of those doses in any trial large enough, long enough, or controlled enough to support the longevity and metabolic claims being made for them.
The muscle question
One of the most-cited reasons to microdose is the fear of losing muscle. A small study reported in 2025 found that up to 40 percent of weight loss from semaglutide came from lean body mass [10]. That number has done serious work in the wellness conversation, and it is at the high end of the published range.
The larger real-world picture is more reassuring. The SEMALEAN study, published in Diabetes, Obesity and Metabolism, followed patients on semaglutide 2.4 mg and reported preserved lean mass and improved muscle function alongside fat loss [11]. Medscape's coverage of related analyses described "sustained weight and fat loss while preserving lean mass and metabolic efficiency" at full dose [12]. A 2025 Endocrine Society press release flagged a real concern, but a more targeted one: women and older adults appear to be at higher risk of muscle loss on these drugs, and higher protein intake appeared to help preserve lean mass in those groups [13].
What microdosing does to that risk is genuinely unknown. If the 40 percent figure holds for you, a smaller dose means a smaller total weight loss, and the proportion lost as lean mass may not improve and could plausibly get worse. The longitudinal studies that would tell us have not been done.
The rebound problem no microdose plan addresses
A BMJ systematic review and a separate Lancet eClinicalMedicine review reached the same conclusion: most patients regain a majority of lost weight within 12 months of stopping a GLP-1 [14][15]. The trials this draws on are STEP-1, STEP-4, and SURMOUNT-4, and the trajectory is consistent.
Microdosing has not been positioned as a maintenance strategy by the manufacturers, because the maintenance studies are still being run. The telehealth programs that do position it that way are extrapolating from absence of evidence rather than presence of it. Anyone starting a microdose program with the goal of "staying on a low dose forever" is committing to an open-ended regimen whose long-term cardiovascular, metabolic, and muscle outcomes are not in the literature.
What to watch
Three things are worth following over the next year. The first is whether Eli Lilly and Novo Nordisk release formal maintenance-dose data that legitimises the lower end of the approved range. The second is whether the FDA, which declared the semaglutide shortage resolved in 2025, tightens the rules on the compounded products that microdose programs rely on. The third is whether the telehealth business model survives: STAT's November 2025 analysis argued that microdosing helps telehealth firms more than it helps patients [2], and the commercial pressure on the brand-name manufacturers is already visible in eMarketer's coverage of their strategic response [16].
None of this means GLP-1 drugs are bad medicine. The pivotal trial data are strong, the weight loss is real, and the cardiovascular benefits extend beyond the scale. What it means is that "less of a good drug" is not a free option. The dose is the intervention, and the intervention has only been studied at full strength.