GLP-1 Microdosing: Evidence, Risks, Chart & Better Alternatives (2026)

Key takeaways

• GLP-1 microdosing (typically 0.05 to 0.25 mg semaglutide weekly) has zero randomized controlled trial support. All weight-loss and cardiovascular data come from standard-dose trials (1.7 to 2.4 mg semaglutide, 10 to 15 mg tirzepatide).
• The FDA-approved starting dose of 0.25 mg semaglutide is already sub-therapeutic, designed for GI adaptation, not weight loss. Clinical weight loss begins at 1.0 mg and peaks at 2.4 mg.
• About 1 in 7 GLP-1 users (14.6%) were microdosing as of late 2025, driven by cost concerns, side-effect avoidance, and social media influence. The trend accelerated during the 2023-2024 shortage.
• Stopping GLP-1 at any dose produces weight regain. A 2025 eClinicalMedicine meta-analysis (18 RCTs, 3,771 participants) found an average regain of 5.63 kg after discontinuation in adults with obesity.
• For full dosing schedules, see GLP-1 weight loss. For side effects, see GLP-1 side effects. For cost, see GLP-1 cost and insurance.

What is GLP-1 microdosing?

GLP-1 microdosing means using a dose of semaglutide or tirzepatide substantially below the FDA-approved therapeutic range. For semaglutide, this typically means 0.05 to 0.25 mg weekly (versus the approved weight-loss maintenance dose of 2.4 mg). For tirzepatide, it means staying at 2.5 mg (the starting dose) long-term or going below it. The goal is to maintain some appetite suppression with fewer side effects and lower cost, rather than pursuing maximum weight loss.

The concept has two distinct use cases that get conflated in online discussions. The first is using low doses as a maintenance strategy after reaching a goal weight on standard doses. The second is using low doses as a primary weight-loss strategy from the start, avoiding the standard titration to higher doses entirely. The evidence for these two approaches is very different. Maintenance microdosing has some logical basis (you need less drug to maintain than to lose). Primary microdosing for weight loss contradicts the dose-response data from every major trial.

Neither use case has been validated in a randomized controlled trial. Everything we know about microdosing comes from case reports, real-world observational data, telehealth provider claims, and social media anecdotes.

Does microdosing GLP-1 actually work?

For significant weight loss: almost certainly not as well as standard dosing, and possibly not at all. The pharmacology is straightforward. GLP-1 receptor agonists produce a dose-dependent effect: higher doses activate more receptors, produce stronger appetite suppression, and result in more weight loss. Every major trial confirms this pattern. In STEP 1, semaglutide at 2.4 mg produced 14.9% weight loss. The lower dose used in Ozempic (2.0 mg, for diabetes) produces roughly 10 to 12%. At 0.25 mg (the starting dose), weight loss is minimal because the dose is designed for tolerability, not efficacy.

One real-world analysis presented at the 2025 European Congress on Obesity tracked 2,700 patients and found that those on approximately 1.0 mg semaglutide achieved about 16% weight loss over 64 weeks, close to the 2.4 mg trial result. This suggested that the dose-response curve may flatten at mid-range doses: you get most of the benefit at 1.0 mg and diminishing returns above that. But 1.0 mg is not a microdose. It is a legitimate therapeutic dose, just below the maximum.

At true microdose levels (0.05 to 0.25 mg), plasma concentrations rarely reach the threshold for meaningful GLP-1 receptor activation. Some individuals may experience mild appetite reduction, but this is difficult to distinguish from placebo effect, especially among highly motivated users who are simultaneously changing their diet and monitoring their intake.

How to microdose GLP-1: the typical approach

Microdosing protocols are not FDA-approved and vary by provider. A common approach: start at 0.125 mg semaglutide weekly (half the standard starting dose), hold for 4 to 8 weeks, and increase to 0.25 mg only if appetite suppression is insufficient. Some providers go as low as 0.05 mg. For tirzepatide, microdosing typically means staying at the 2.5 mg starting dose indefinitely rather than escalating.

Because standard injectable pens (Wegovy, Ozempic) deliver fixed doses and cannot be easily subdivided, microdosing typically requires compounded semaglutide from a compounding pharmacy. This introduces additional variables: compounded drug quality is not FDA-verified, concentrations can vary between pharmacies, and the regulatory landscape for compounded GLP-1 is tightening as the drug shortage has ended.

Any microdosing approach should be supervised by a licensed healthcare provider who can monitor your response, adjust dosing, and watch for complications. Self-adjusting GLP-1 doses based on social media guidance is risky.

Why people microdose GLP-1

Three drivers: cost, side effects, and the drug shortage legacy. At full dose, brand-name GLP-1 costs $1,000+ per month without insurance. Microdosing with compounded semaglutide can cut the medication cost to $50 to $100 per month. GI side effects (nausea, vomiting, diarrhea) are dose-dependent; lower doses mean fewer side effects. And during the 2023-2024 shortage, patients who rationed their supply discovered that spaced-out or reduced doses still provided some appetite suppression.

Social media accelerated the trend. In October 2024, a widely shared podcast interview promoted microdosing as a "gamechanger." By late 2025, an Evidation survey found that roughly 14.6% of GLP-1 users, about 1 in 7, were microdosing. The appeal is understandable: cheaper, gentler, and seemingly effective enough. The problem is that "seemingly" is doing a lot of work in that sentence. Without controlled data, we cannot separate genuine pharmacological effect from placebo, behavioral change, and selection bias (people who microdose are often highly motivated and health-conscious).

What happens when you stop microdosing?

Weight regain, just like stopping at any dose. A 2025 systematic review and meta-analysis in eClinicalMedicine (Tzang et al., 18 RCTs, 3,771 participants) found that adults with obesity who stopped GLP-1 receptor agonists regained an average of 5.63 kg, with HbA1c rising 0.25 percentage points and waist circumference, BMI, systolic blood pressure, and fasting glucose all deteriorating. Subgroup analysis showed greater regain with semaglutide (8.21 kg) than liraglutide (4.29 kg). Pooled data on tirzepatide discontinuation shows a comparable trajectory, with most patients regaining a substantial share of their loss within a year.

This is the strongest argument against microdosing as a cost-saving strategy. If you microdose, lose less weight, and then stop, you regain what you lost. You spent less money but also achieved less result. Standard dosing followed by a thoughtfully managed dose reduction for maintenance may ultimately be more cost-effective than starting and staying at an ineffective dose.

The maintenance question is where microdosing has its most legitimate application. After reaching a goal weight on standard doses, transitioning to a lower maintenance dose makes pharmacological sense. Some providers report success with patients maintaining weight on 0.5 to 1.0 mg semaglutide after achieving their goal on 2.4 mg. But this is maintenance microdosing, not primary microdosing, and the evidence is still anecdotal.

GLP-1 microdosing chart: semaglutide and tirzepatide

No standardized microdosing chart exists because microdosing is off-label. That said, the doses circulating in telehealth and social media communities typically follow this pattern:

These charts are shared widely on Reddit, TikTok, and telehealth provider websites, but none are backed by clinical trial data. The standard titration schedule (right column) is the only evidence-based approach. The microdose columns represent what patients and some providers are doing in practice without regulatory or clinical trial validation.

Is GLP-1 microdosing safe?

Probably, at the drug level: lower doses generally mean fewer side effects. The safety concerns are indirect. First, microdosing almost always requires compounded medication, which is not FDA-approved and has quality variability. The FDA has received a substantial volume of adverse event reports tied to compounded semaglutide and tirzepatide, and has cited manufacturing and quality problems in its enforcement actions against compounders. Second, inadequate dosing may delay treatment of obesity-related comorbidities (diabetes, cardiovascular disease) that standard-dose GLP-1 effectively manages. Third, if microdosing proves ineffective and the patient stops, they may lose motivation to try standard dosing later.

The compounding risk is real. When you microdose using a compounding pharmacy, you are relying on a non-FDA-verified product with custom concentrations. Dosing errors, contamination, and inconsistent potency are documented problems. The FDA has taken enforcement action against multiple compounding pharmacies for GLP-1 product quality issues. For side effect information at any dose, see our GLP-1 side effects guide.

How much does microdosing save?

Compounded microdose semaglutide costs roughly $50 to $150 per month through telehealth platforms, compared to $199 to $449 for FDA-approved medication through manufacturer cash-pay programs. For insured patients using savings cards ($25 per month), microdosing offers no cost advantage because you are already at minimum cost. The savings are meaningful only for uninsured patients paying cash, and they must be weighed against the reduced efficacy and the regulatory uncertainty around compounded GLP-1 products.

With the Wegovy pill now available through NovoCare at $149 per month (starter dose), the cost gap between FDA-approved medication and compounded microdosing has narrowed substantially. For a full cost comparison, see our GLP-1 cost and insurance guide.

Microdosing for GLP-1 maintenance after weight loss

This is the most evidence-adjacent use of microdosing. After reaching a goal weight on standard doses, some providers transition patients to a lower maintenance dose (0.5 to 1.0 mg semaglutide, or 2.5 to 5.0 mg tirzepatide) to sustain appetite suppression at reduced cost and side-effect burden. A 2025 case series reported that patients dosing semaglutide every 2 to 4 weeks (rather than weekly) maintained most of their weight loss.

The logic is sound: you need less drug to maintain a weight than to lose it, because the caloric deficit required for maintenance is smaller. But the STEP 1 extension data is cautionary. When patients stopped semaglutide entirely after 68 weeks, they regained two-thirds of lost weight within a year. That regain started immediately, not gradually. If a maintenance microdose is too low to sustain appetite suppression, the result will be the same as stopping entirely, just slower.

The safest approach to maintenance dosing is to work with your prescriber to find the lowest effective dose, not the lowest dose you can find on the internet. "Effective" means: your appetite stays manageable, your weight remains within 3 to 5% of your goal, and your metabolic markers (blood sugar, blood pressure, lipids) stay stable.

The compounding crackdown: what it means for microdosing

Microdosing depends almost entirely on compounded semaglutide, because commercial pens deliver fixed doses that cannot be safely subdivided below their calibrated increments. That supply chain is closing. The FDA declared the tirzepatide shortage resolved in December 2024 and the semaglutide shortage resolved in February 2025, ending the legal exemption that permitted mass-scale compounding. On April 30, 2026, the agency proposed removing semaglutide, tirzepatide, and liraglutide from the 503B bulks list entirely, which would permanently bar outsourcing facilities from compounding them even during future shortages.

The timeline matters for anyone currently microdosing. The FDA set staged transition deadlines: 503A pharmacies had to stop compounding tirzepatide by mid-February 2025 and semaglutide by April 22, 2025, with 503B outsourcing facilities following in March and May 2025 respectively. The Outsourcing Facilities Association sued to block the delisting, but a federal district court denied its preliminary injunction motion in April 2025, and the deadlines held.

A narrow pathway remains. Section 503A patient-specific compounding operates under a different legal framework that does not depend on shortage-list status. A state-licensed pharmacy can still compound a customized dose for an individual patient with a valid prescription, provided it is not producing "essentially a copy" of a commercially available drug regularly or in inordinate amounts. This is the crack through which most microdosing prescriptions currently flow. Whether that crack stays open depends on how the FDA interprets and enforces the "essentially a copy" standard going forward.

In March 2026, the FDA issued more than 50 warning letters to telehealth firms and compounding pharmacies in a single enforcement wave, the largest coordinated action against GLP-1 compounders in US history. Violations cited included compounding after shortage resolution, misleading marketing, and improper manufacturing practices. If you are on a microdosing protocol supplied by a telehealth platform, it is worth asking your provider directly which pharmacy fills your prescription and under what regulatory pathway.

Why commercial pens make microdosing difficult

Wegovy and Ozempic pens are calibrated to deliver specific doses (0.25, 0.5, 1.0, 1.7, 2.4 mg) and lock at those increments. There is no dial that goes lower. Some patients attempt to microdose by drawing partial doses from a pen or vial with a separate syringe, but this introduces dosing errors, sterility risks, and voids the device's calibration guarantees. Novo Nordisk and Eli Lilly both explicitly warn against this practice.

The pharmacokinetics compound the problem. Semaglutide has a half-life of approximately seven days, which means plasma levels accumulate over four to five weeks before reaching steady state. A microdose administered weekly builds to a steady-state concentration proportional to the dose. At 0.125 mg weekly, steady-state plasma semaglutide sits roughly one-twentieth of what the 2.4 mg maintenance dose produces. Receptor occupancy at that concentration is minimal.

Some patients instead space standard doses further apart, taking 0.25 mg every two or three weeks rather than a smaller dose weekly. This produces peaks and troughs rather than a stable plateau. Whether intermittent partial receptor activation delivers meaningful appetite suppression is an open pharmacological question with no trial data behind it. Anecdotally, some patients report that appetite returns noticeably in the days before their next spaced-out dose, which is consistent with concentrations falling below the effective threshold.

Who might actually benefit from a lower GLP-1 dose?

There are legitimate clinical scenarios for using less than the maximum dose, and they should not be confused with microdosing. Patients who cannot tolerate dose escalation because of persistent nausea, patients who reach their weight goal before hitting the maximum dose, patients transitioning to maintenance after successful loss, and patients with a history of severe GI sensitivity all have reasons to stay at a lower dose. None of these are microdosing in the 0.05 to 0.25 mg sense.

The distinction is between a lower therapeutic dose and a sub-therapeutic dose. Semaglutide at 1.0 mg weekly is a lower therapeutic dose with real evidence behind it: the real-world analysis presented at the 2025 European Congress on Obesity found approximately 16% weight loss at that level over 64 weeks. Semaglutide at 0.125 mg is a sub-therapeutic dose with no evidence at all. Both get called "microdosing" in online discussions, and that conflation is the source of most of the confusion in this space.

If your goal is fewer side effects, the evidence-based approach is a slower titration, not a permanently lower dose. Extending each titration step from four weeks to six or eight weeks gives your GI system more time to adapt while still eventually reaching a therapeutic level. Prescribers do this routinely for patients with tolerability problems, and it does not sacrifice the eventual outcome.

Microdosing and side effects: is it actually gentler?

Yes, lower doses produce fewer gastrointestinal side effects. That relationship is well established and dose-dependent across every GLP-1 trial. Nausea, vomiting, diarrhea, and constipation all increase in frequency and severity as the dose escalates. At 0.25 mg semaglutide, roughly 15 to 20% of patients report nausea. At 2.4 mg, that figure climbs to around 44%.

But the tradeoff is direct. The same receptor activation that slows gastric emptying enough to cause nausea is the mechanism that produces satiety and reduces caloric intake. You cannot separate the therapeutic effect from the side effect profile, because they share a mechanism. A dose gentle enough to eliminate nausea entirely is usually a dose too low to meaningfully suppress appetite.

There is a middle path that most obesity medicine specialists actually recommend: reach a therapeutic dose slowly, manage side effects actively (smaller meals, avoiding high-fat foods, staying hydrated, anti-nausea medication if needed), and accept that the worst symptoms are typically transient. GI side effects peak within the first two to four weeks after each dose increase and fade as the body adapts. Patients who push through the adaptation period usually arrive at a therapeutic dose with tolerable symptoms. For a full breakdown of what to expect and how to manage it, see our GLP-1 side effects guide.

Tirzepatide microdosing: what is different

Tirzepatide microdosing usually means staying at 2.5 mg weekly indefinitely rather than escalating to 5, 7.5, 10, or 15 mg. This is a different situation from semaglutide microdosing because 2.5 mg is the FDA-approved starting dose, not a sub-therapeutic fraction of it. Some weight loss does occur at 2.5 mg, though substantially less than at maintenance doses.

In SURMOUNT-1, the 5 mg dose produced 15.0% weight loss, 10 mg produced 19.5%, and 15 mg produced 20.9% at 72 weeks. The trial did not include a 2.5 mg maintenance arm because 2.5 mg is designed as a four-week titration step, not a therapeutic target. Patients who stay at 2.5 mg are operating outside the studied protocol, though closer to the evidence base than someone taking 0.125 mg of semaglutide.

Tirzepatide also has a shorter half-life than semaglutide, roughly five days versus seven, which means plasma levels fall faster between doses and steady state is reached slightly sooner. Practically, this makes tirzepatide slightly less forgiving of missed or delayed doses. For dosing schedules and the full clinical comparison, see our GLP-1 weight loss guide.

Does microdosing preserve muscle better?

This is a common claim in microdosing communities, and the logic is plausible but unproven. The argument: slower weight loss allows more time for the body to adapt, and a smaller caloric deficit produces proportionally less lean-mass loss. The DXA substudy in STEP 1 found that roughly 40% of weight lost on semaglutide was lean mass, a figure that has driven considerable concern about sarcopenia risk.

No study has compared body composition outcomes between microdosing and standard dosing, so the muscle-preservation claim rests on inference rather than data. What is well established is that resistance training and adequate protein intake preserve lean mass during any weight loss, regardless of the rate or the mechanism driving the deficit. A patient on a standard dose who lifts weights and eats 1.0 to 1.2 grams of protein per kilogram of body weight will almost certainly preserve more muscle than a microdosing patient who does neither.

If muscle preservation is your primary concern, the intervention with actual evidence behind it is resistance training, not dose reduction. For nutritional strategies during GLP-1 treatment, see our best GLP-1 supplements guide.

Low-dose GLP-1 for non-weight-loss purposes

A separate and more evidence-supported use of low-dose GLP-1 involves indications other than weight loss. A 2025 randomized controlled trial published in JAMA Psychiatry tested low-dose semaglutide (0.25 to 0.5 mg weekly) for alcohol use disorder and found reduced drinking. Those doses are low relative to weight-loss protocols but sit within the FDA-approved titration range, not below it. The larger Lancet trial that followed used the full 2.4 mg dose; we cover what the GLP-1 and alcohol research actually found in detail.

Interest is also growing in low-dose GLP-1 for metabolic markers independent of weight: fasting glucose, insulin sensitivity, and inflammatory markers. Some of these effects appear at doses lower than those required for substantial weight loss, which makes pharmacological sense because different downstream pathways may have different receptor-occupancy thresholds. This is an active research area with promising early signals and no definitive answers.

What has not been demonstrated at low doses: the cardiovascular benefit. The 20% reduction in major adverse cardiovascular events shown in the SELECT trial came from semaglutide 2.4 mg, the maximum weight-management dose. Whether a fraction of that dose delivers a fraction of that benefit, or no benefit at all, is unknown. Patients choosing microdosing specifically to capture cardiometabolic protection should understand that they are extrapolating well beyond the evidence.

How to talk to your prescriber about microdosing

Be direct about your reasons. Prescribers hear the microdosing question constantly and generally respond better to a stated concern (cost, side effects, fear of losing too much weight) than to a request for a specific protocol found online. Framing it as "I am struggling with nausea and want to discuss options" opens a clinical conversation. Framing it as "I want to microdose at 0.125 mg" positions your prescriber to either accept an unvalidated protocol or refuse.

Useful questions to bring: What is the lowest dose that is likely to work for my situation? Can we extend the titration schedule instead of stopping at a low dose? If cost is the barrier, are there manufacturer programs or formulary alternatives I have not tried? If I reach my goal weight, what does a maintenance plan look like? These questions get at the underlying problem rather than jumping to a specific solution.

If your prescriber declines to support microdosing, the reason is usually that they cannot point to evidence supporting it and are unwilling to prescribe outside the studied range without justification. That is a defensible clinical position, not obstruction. Many prescribers will, though work with you on a slower titration, a lower maintenance dose after successful weight loss, or a different medication with a better tolerability profile for you.

Cheaper alternatives to microdosing

If cost is the reason you are considering microdosing, several FDA-approved options now cost less than compounded microdose protocols did a year ago. The Wegovy oral tablet through NovoCare starts at $149 per month for the starter dose. Wegovy injection through NovoCare runs $199 to $349. Zepbound through LillyDirect costs $299 to $449. All are FDA-approved products at therapeutic doses.

For insured patients, manufacturer savings cards bring the cost to roughly $25 per fill, which no compounded microdose protocol can beat. If your insurance denied coverage, appealing the denial is often more productive than switching to an unvalidated dosing strategy. Denial overturn rates have improved as the clinical evidence base has strengthened.

Medicare beneficiaries gained a new pathway in July 2026 through the GLP-1 Bridge program, which provides FDA-approved GLP-1 medications for obesity at a $50 monthly copay. For the full cost landscape and strategies for reducing out-of-pocket expense, see our GLP-1 cost and insurance guide.

The bottom line on GLP-1 microdosing

Microdosing occupies a strange position: widely practiced, heavily discussed, and completely unstudied. Roughly one in seven GLP-1 users report doing it, yet not a single randomized controlled trial has evaluated it as a weight-loss strategy. Everything we know comes from anecdote, observational data, and pharmacological inference.

What the pharmacology suggests: doses below approximately 0.5 mg semaglutide are unlikely to produce clinically meaningful weight loss, because receptor occupancy at those plasma concentrations falls below the threshold that drives appetite suppression in trials. What real-world data suggests: mid-range doses around 1.0 mg may capture most of the benefit of the maximum dose, which is a genuinely useful finding but describes lower therapeutic dosing, not microdosing.

What is clear: stopping GLP-1 treatment at any dose produces weight regain. The 2025 eClinicalMedicine meta-analysis (Tzang et al.) documented an average 5.63 kg regain in adults with obesity, with greater regain on semaglutide (8.21 kg) than liraglutide (4.29 kg) and with longer follow-up (7.31 kg beyond 26 weeks vs 2.51 kg). A separate Obesity Reviews systematic review (Berg et al., 2025) reached consistent conclusions on body habitus changes after stopping. Pooled analyses of tirzepatide discontinuation show a similar pattern: most patients regain a substantial share of their weight loss within a year of stopping. Obesity behaves like a chronic condition, and treating it at a dose too low to work is not a cost-saving strategy. It is paying for something that does not deliver.

If you are considering microdosing, the conversation worth having with your prescriber is not "what is the smallest dose I can take" but "what is the smallest dose that will actually work for me, and how do we get there with tolerable side effects." Those are different questions with very different answers.

Microdosing and the 2024 social media origin story

The GLP-1 microdosing trend has a traceable origin. When Wegovy and Mounjaro spent stretches of 2023 and 2024 on the FDA drug shortage list, compounding pharmacies could legally produce copies, and patients rationing scarce supply discovered that doses spaced two or three weeks apart still blunted their appetite. The habit outlived the shortage. In October 2024, a widely circulated podcast interview with a performance physician described microdosing as a "gamechanger," and the framing spread rapidly through wellness media and influencer channels.

By late 2025, an Evidation survey found that roughly 14.6% of GLP-1 users, about one in seven, reported microdosing. The demographic skewed toward women aged 35 to 60, and marketing increasingly targeted perimenopausal women with language about "personalized" and "gentle" dosing. STAT News documented how the concept moved from patient workaround to commercial product category within roughly a year.

Understanding this history matters because it explains why the evidence base is thin. Microdosing did not emerge from a clinical hypothesis that researchers then tested. It emerged from supply constraints and cost pressure, spread through social proof, and was then commercialized by telehealth platforms that had compounding relationships already in place. The scientific question of whether sub-therapeutic GLP-1 dosing produces meaningful outcomes has never been formally asked in a controlled setting, because the practice arrived through a channel that does not generate trials.

That does not make microdosing wrong. Plenty of clinical practices began as observation and were later validated. It does mean that anyone microdosing today is participating in an uncontrolled experiment with no safety monitoring, no efficacy endpoints, and no comparison group. If it works for you, there is no mechanism to record that. If it does not, there is no mechanism to record that either.

Frequently asked questions

What is microdosing GLP-1?

Using a semaglutide or tirzepatide dose well below the FDA-approved therapeutic range (0.05 to 0.25 mg semaglutide weekly versus the approved 1.7 to 2.4 mg). It is an off-label practice with no clinical trial support.

Does low dose GLP-1 work?

At true microdose levels (under 0.5 mg semaglutide), no clinical trial has demonstrated meaningful weight loss. At mid-range doses (1.0 mg), real-world data suggests weight loss approaching what the maximum dose achieves. "Low dose" and "microdose" are not the same thing.

How to microdose GLP-1?

Only under medical supervision. Microdosing requires a prescription, typically for compounded semaglutide at a custom concentration. Your prescriber determines the starting dose and monitors your response. Self-adjusting GLP-1 doses is unsafe.

Is GLP-1 microdosing FDA-approved?

No. The FDA has approved semaglutide (Wegovy) at 2.4 mg weekly and tirzepatide (Zepbound) at up to 15 mg weekly for weight management. Doses below the approved range are considered off-label use.

Watch: GLP-1 dosing explained

Published research cited in this article

This statement has not been evaluated by the Food and Drug Administration. This product is not intended to diagnose, treat, cure, or prevent any disease. This article is for informational purposes only and is not intended as medical advice. GLP-1 microdosing is an off-label practice not validated by clinical trials. Do not adjust GLP-1 medication doses without consulting your healthcare provider. Consult your healthcare provider before starting, stopping, or changing any GLP-1 treatment.